Professional Documents
Culture Documents
Edited by
gretchen helmke
steven levitsky
A catalog record for this book is available from the British Library.
Contents
Preface vii
Introduction 1
gretchen helmke and steven levitsky
pa r t i t h e i n f o r m a l p o l i t i c s o f
executive-legislative relations
1 Accommodating Informal Institutions and Chilean Democracy 33
peter siavelis
pa r t i i i i n f o r m a l i n s t i t u t i o n s a n d pa r t y p o l i t i c s
7 The Birth and Transformation of the Dedazo in Mexico 143
joy langston
Conclusion 274
gretchen helmke and steven levitsky
Afterword
On Informal Institutions, Once Again 285
guillermo o’donnell
Notes 291
References 313
List of Contributors 337
Index 341
Preface
Political reality can be compelling. The sweeping regime changes of the 1980s
and 1990s brought democratic institutions to virtually every country in Latin Amer-
ica, but the quality or performance of those institutions has disappointed both schol-
ars and policymakers alike. Military coups have largely disappeared, but presidents
continue to be forced from office before the end of their mandate; constitutional
liberties have been restored, but security forces kill some citizens with impunity;
legislators seem more interested in making money than in making policy; corruption
and clientelism remain widespread, and in many rural (and some urban) areas, the
rule of law effectively does not exist. This book contends that in order to understand
how—and how well—democratic institutions work in Latin America, scholars must
go beyond the study of formal institutions and take seriously informal ‘‘rules of the
game.’’ The book presents a conceptual and theoretical framework for analyzing
how formal and informal institutions interact in new democracies. Although it
focuses on Latin America, its lessons are broadly applicable throughout the develop-
ing and postcommunist worlds.
The idea for this volume emerged out of a series of conversations that began
nearly a decade ago. As researchers beginning fieldwork in Argentina during the
mid-1990s, we were struck by the vast gap between the formal institutions we had
come to study (political parties in one case, courts in the other) and the informal
realities we encountered on the ground. Our respective efforts to make sense of these
patterns were heavily influenced by the work and teaching of Guillermo O’Donnell.
O’Donnell, who has written an afterword to this volume, is a major intellectual
inspiration behind it.
We have incurred many debts in bringing this project to fruition. The volume
emerged out of two conferences on informal institutions. The first, entitled ‘‘Infor-
mal Institutions and Politics in the Developing World,’’ was held at Harvard Univer-
sity in April 2002. It was generously funded by the Weatherhead Center for Inter-
national Affairs and the David Rockefeller Center for Latin American Studies. We
viii Preface
are particularly grateful for the support of Weatherhead Center director Jorge Do-
mínguez, and for the dedicated organizational work of Jeana Flahive. Our own
conceptual and theoretical ideas about informal institutions were heavily informed
by this conference, and we thank participants Kathleen Collins, Keith Darden, Jorge
Domínguez, Dennis Galvan, Robert Gay, Kathryn Hendley, Jim Johnson, Jack
Knight, Hans-Joachim Lauth, Melanie Manion, José Luis Medina, María Victoria
Murillo, Andreas Schedler, Rudra Sil, Lily Tsai, and Lucan Way.
The second conference, entitled ‘‘Informal Institutions and Politics in Latin
America: Understanding the Rules of the Game,’’ was held at the Kellogg Institute
for International Studies, University of Notre Dame, in April 2003. The conference
was generously supported by the Kellogg Institute, through a grant from the Coca
Cola Foundation. We are particularly thankful for the support of Frances Hagopian
and Scott Mainwaring, as well as Christopher Welna, Holly Rivers, and Dawn
Dinovo, and to conference participants Rebecca Bill Chavez, Martín Böhmer, Jorge
Buendia, Michael Coppedge, Jack Knight, Miriam Kornblith, Susan Stokes, Ignacio
Walker, and Kurt Weyland for their insightful comments.
In addition, we have received extremely useful comments along the way from
Jorge Domínguez, Anna Grzymala-Busse, Peter Hall, Goran Hyden, Lisa Mar-
tin, María Victoria Murillo, Shannon O’Neil Trowbridge, Benjamin Smith, Hillel
Soifer, Lucan Way, and Jason Wittenberg. Maria Koinova, Elena Plaxina, and Hillel
Soifer provided critical research and editorial assistance. We also thank the staff at
the Johns Hopkins University Press, and particularly Henry Tom, for their careful
assistance in bringing the book to press.
Excerpts from the Introduction are taken from ‘‘Informal Institutions and Com-
parative Politics: A Research Agenda,’’ by Gretchen Helmke and Steven Levitsky, in
Perspectives on Politics, vol. 2, no. 4, December 2004, pp. 725–740. Copyright ∫ 2004
by the American Political Science Association. Reprinted with the permission of
Cambridge University Press.
Finally, Steve Levitsky thanks his wife, Liz Mineo, and daughter, Alejandra Sol
Mineo-Levitsky, for keeping him focused on the rules of the game that really matter.
Gretchen Helmke thanks Mitch Sanders and her father, Stephen Helmke, for their
support and encouragement.
Informal Institutions and Democracy
This page intentionally left blank
Introduction
Over the past two decades, a scholarly consensus has emerged around the central-
ity of political institutions. In Latin America, recent research on executive-legislative
relations, electoral and party systems, judicial politics, bureaucracies, and federalism
has shed new light on how institutional design affects the stability and quality of
democracy.∞ Nevertheless, persistent problems of corruption, clientelism, executive-
legislative conflict, and the ‘‘unrule of law’’ cast doubt on whether an exclusive focus
on ‘‘parchment’’≤ institutions is sufficient for understanding what drives politics in
the region (O’Donnell 1996a, 1999c; Weyland 2002a). Scholars such as Guillermo
O’Donnell and Douglass North have argued that informal institutions—or rules and
procedures that are created, communicated, and enforced outside the officially
sanctioned channels—are often as important as their formal counterparts in structur-
ing the ‘‘rules of the game.’’
Informal rules coexist with formal democratic institutions throughout Latin
America. In Mexico during much of the twentieth century, presidents were selected
not according to rules laid out in the constitution, the electoral law, or the statutes of
the governing Party of the Institutional Revolution (PRI), but rather by means of the
dedazo (‘‘finger tapping’’), an unwritten code that gave the sitting president the right
to choose his successor, specified the candidate pool, and prohibited potential candi-
dates from openly seeking the job. In Chile, notwithstanding a constitution that
created one of the most powerful presidencies in the world, informal power-sharing
arrangements in place since democratization have induced presidents to systemati-
cally underutilize that power. As a result, Chile is viewed as an exception in a re-
2 gretchen helmke and steven levitsky
≤ In the conceptual realm, they address the question of what informal institu-
tion are and, crucially, what they are not.
≤ In the theoretical realm, the chapters explore four central questions: (1) What
are the distinct ways in which formal and informal institutions interact?
(2) What are the effects of informal institutions, particularly with respect to
the quality, performance, and stability of democracy? (3) What are the origins
4 gretchen helmke and steven levitsky
This introductory chapter provides an overview of these questions. The first section
examines the concept of informal institution. It makes the case for a narrow defini-
tion that permits scholars to distinguish between informal institutions and other
informal phenomena, such as weak institutions, culture, and non-rule-bound be-
havior patterns. In the second section we provide an overview of how informal
institutions affect the performance and quality of democracy in Latin America. The
third section develops a typology of formal-informal institutional interaction. Ex-
panding on the work of Hans-Joachim Lauth (2000), we distinguish among four
types of informal institution: complementary, accommodating, competing, and sub-
stitutive. The fourth section explores a set of theoretical issues at the frontier of
research on informal institutions, including the critical, but underresearched, ques-
tions of informal institutional emergence and change. Finally, we discuss some
challenges related to research on informal institutions, such as issues of identifica-
tion, measurement, and comparison.
Informal institutions merit our attention because they shape how democracy
works—for both good and ill. Perhaps not surprisingly, much of the existing literature
on informal institutions in new democracies focuses on their negative effects. In
particular, studies have highlighted ways in which corruption, clientelism, and patri-
monialism undermine the effectiveness of democratic, state, and market institu-
tions.∞≤ The chapters in this volume tell a more mixed story. Although several essays
provide systematic evidence of how informal institutions erode the quality of demo-
cratic institutions, others point to the ambiguous, double-edged, and even positive
effects of informal rules. Particularly where formal state and regime institutions are
weak, ineffective, or insufficiently democratic, informal rules may enhance the
performance and stability of democracy. In this section we examine the effects of
informal institutions in four key areas of democratic politics: representation, ac-
countability, governability, and citizenship and the rule of law.
Political Representation
Several of the essays in this volume grapple with issues of political representation.
Evidence of a growing gap between citizens and politicians in Latin America is
abundant: it includes declining party identification and voter turnout (Hagopian
1998, 114–21), high levels of electoral volatility (Mainwaring and Scully 1995; Roberts
and Wibbels 1999), the rise of personalistic or ‘‘neopopulist’’ outsiders (Roberts 1995;
Weyland 1996, 1999), and, in a few countries, large-scale protest against the entire
political elite. Scholars have linked this ‘‘representation gap’’ to informal institutions
such as clientelism, patrimonialism, and corruption (Fox 1994; O’Donnell 1996b;
Mainwaring 1999; Brusco et al. 2004). These particularistic institutions are thus said
to erode or prevent the establishment of programmatic linkages between parties and
citizens.
The chapters in this volume find new evidence in support of these claims. For
example, in his comparative analysis of state-level legislative politics in Brazil, Scott
Desposato finds that clientelism erodes legislative parties’ capacity to represent vot-
ers on programmatic issues. In the highly clientelistic state of Piauí, he finds that
party discipline is low, individual legislators rarely take public positions on issues,
introduction 9
Democratic Accountability
Recent evidence from Latin America suggests that democratic elections are often
insufficient mechanisms for ensuring government accountability and responsive-
ness. Using O’Donnell’s terms (1994), accountability has been found wanting in two
key areas: vertical accountability, or the degree to which citizens are able to reward or
punish officials for their performance in office (O’Donnell 1994; Stokes 2001), and
horizontal accountability, or the degree to which public officials are responsible to
(or checked by) other agencies and institutions of the state (O’Donnell 1994, 1999b;
Schedler et al. 1999; Mainwaring and Welna 2003). In much of the region, citizens
and their representatives are said to lack effective mechanisms to oversee and, when
necessary, punish officeholders who abuse power. As a result, elected officials rou-
tinely betray their mandates, abuse their authority, and ignore constituents’ demands.
Here, too, informal institutions are widely viewed as obstacles to normatively
desirable outcomes, and for good reason. Because they are unwritten and unregu-
lated, informal rules generally lack the transparency or public oversight that is often
essential to accountability. It is difficult to use the law or public agencies to hold a
politician accountable for breaking rules that—by definition—are not on the books.
Several chapters in this volume explore the link between the nontransparent nature
of informal institutions and deficits of accountability. For example, in their essay on
informal party organization, Freidenberg and Levitsky argue that informal finance,
10 gretchen helmke and steven levitsky
decision-making, and other intraparty processes widen the gap between parties’
public faces and their ‘‘real’’ power structures, which limits the capacity of activists
and voters to hold party leaders accountable. In line with this analysis, Langston
argues in her chapter that PRI elites preferred to govern through informal institu-
tions such as the dedazo (rather than create formal authoritarian rules) in part
because they helped prevent local party activists from holding them accountable.
Had the dedazo system been written into PRI statutes or the constitution, it would
have been open to public scrutiny—and thus more vulnerable to public challenge.
Informal institutions need not always undermine accountability, however. As
Stokes’s chapter shows, informal norms may also reinforce or sustain formal institu-
tions of accountability. For elections to function effectively as mechanisms of verti-
cal accountability, Stokes argues, certain shared expectations about how citizens will
evaluate politicians are essential. Only when it is widely believed that citizens will
follow an informal decision rule to vote retrospectively, weighing a politician’s past
performance in deciding how to vote (as opposed to, say, exchanging their votes for
particularistic favors), will politicians act responsively and citizens expect such re-
sponsiveness. Drawing on a comparison of cities and provinces in Argentina, Stokes
finds that democratic institutions work better where such shared expectations exist
(Mar del Plata) than where they do not (Misiones).
Of course, informal institutions may also ensure accountability in less appealing
ways. As Samuels shows in his analysis of campaign finance in Brazil, particularistic
relationships are critical to sustaining illicit campaign finance contracts, for they
help overcome the credible commitment problems inherent in illegal transactions.
By providing trust, familiarity, reputations, and repeated interaction, particularistic
norms help business leaders and politicians hold each other accountable in a con-
text in which no legal recourse is available. Hence, they serve as the ‘‘glue’’ that
sustains informal campaign finance contracts. Todd Eisenstadt’s chapter on infor-
mal mechanisms of postelectoral conflict resolution (concertacesiones) also finds an
ambiguous effect with respect to accountability. On the one hand, negotiated agree-
ments between Mexico’s ruling PRI and the opposition National Action Party (PAN)
brought a degree of vertical accountability in that they removed from power ruling
party candidates who had won elections through fraud. On the other hand, choosing
mayors and governors though backroom bargaining completely severed the (how-
ever fictitious) link between the voting process and electoral outcomes, which can
hardly be said to enhance vertical accountability.
Taken together, these chapters thus show that informal rules may provide the
bases for credible commitment and some degree of accountability. However, as
Samuels points out, they are less likely to provide the ‘‘right’’ kind of credible
introduction 11
commitments (i.e., those that enhance public accountability) needed for demo-
cratic consolidation.
Democratic Governance
served as the basis for one of the few institutionalized nondemocratic mechanisms of
succession in the modern world. Similarly, Eisenstadt’s essay suggests that the PRI’s
use of concertacesiones may have helped prevent large-scale political conflict during
the 1990s. Although these informal mechanisms of governability probably slowed
Mexico’s democratic transition, they may also have contributed to the relative sta-
bility of that transition.
As the previous section makes clear, formal and informal institutions interact in
diverse ways, with diverse consequences. To make sense of these various patterns, in
this section we develop a typology of formal-informal institutional relationships. The
typology is based on two dimensions.∞∑ The first is the degree of convergence be-
tween formal and informal institutional outcomes. The distinction here is whether
following the informal rules produces a result substantively similar to or different
from that expected from a strict and exclusive adherence to the formal rules. Where
following the informal rule leads to a substantively different outcome, formal and
informal institutions may be said to diverge. Where the two outcomes are not sub-
stantively different, formal and informal institutions converge. The second dimen-
sion is that of the effectiveness of the relevant formal institutions. By effectiveness we
mean the extent to which rules and procedures that exist on paper are enforced or
complied with in practice.∞∏ Where formal institutions are effective, actors believe
there is a high probability that noncompliance will be sanctioned by official authori-
ties. Where formal rules and procedures are ineffective, actors believe the proba-
bility of enforcement (and hence the expected cost of violation) to be low. These two
dimensions produce the fourfold typology shown in figure I.1.
Accommodating Informal Institutions. The lower left cell of figure I.1, which com-
bines effective formal institutions and divergent outcomes, corresponds to what we
call accommodating informal institutions. These informal institutions create incen-
tives to behave in ways that alter the substantive effects of formal rules, but without
directly violating them. In other words, they contradict the spirit, but not the letter, of
the formal rules. Accommodating informal institutions are often created by actors
who dislike outcomes generated by the formal rules but are unable to change or
openly violate those rules. As such, these institutions often help reconcile these
actors’ interests with the existing formal institutional arrangements. A classic exam-
ple is Dutch consociationalism, a set of ‘‘informal, unwritten rules’’ of cross-party
accommodation and power-sharing that included extensive consultation in policy-
making, mutual vetoes, and a proportional distribution of government jobs (Lijphart
1975, 122–38). Although consociational arrangements violated the democratic spirit
of the Dutch constitution by limiting the power of the vote, they enhanced regime
stability by dampening class and religious conflict (Lijphart 1975, 137–38).
Siavelis’s contribution to this volume offers another example of an accommodat-
ing informal institution. According to Siavelis, the 1980 Chilean constitution was
‘‘among the least conducive to effective democracy in Latin America.’’ A powerful
presidency and majoritarian electoral rules ‘‘create disincentives for cooperation,
coalition formation and political accommodation,’’ which threatened the quality, if
not the stability, of Chile’s post-1989 democracy. Lacking the political strength to
abolish or reform the Pinochet-era constitution, elites within the governing demo-
cratic Concertación developed a set of informal procedures that counteracted its
effects. Informal mechanisms such as the cuoteo, partido transversal, and demo-
cracia de los acuerdos created incentives for interparty and interbranch cooperation
and consultation, which ‘‘mitigated the most negative characteristics of exaggerated
presidentialism.’’
Competing Informal Institutions. To the right side of figure I.1 are instances of
informal institutions that coexist with ineffective formal institutions. The lower right
cell combines ineffective formal rules and divergent outcomes, producing compet-
ing informal institutions. These informal institutions structure incentives in ways
that are incompatible with the formal rules: to follow one rule, actors must violate
another. Competing informal institutions trump their formal counterparts, generat-
ing outcomes that diverge markedly from what is expected from the formal rules. An
example is systemic corruption. In postwar Italy, norms of corruption were ‘‘more
powerful than the laws of the state: the latter could be violated with impunity, while
anyone who challenged the conventions of the illicit market would meet with cer-
16 gretchen helmke and steven levitsky
tain punishment’’ (Della Porta and Vannucci 1999, 15). Similarly, particularistic
norms such as clientelism and patrimonialism are often said to subvert formal state,
market, and electoral institutions (O’Donnell 1996b; Borozc 2000; Lauth 2000).
Brinks’s chapter on police violence offers an example of a competing informal
institution. Although Brazilian law prohibits police officers from killing suspected
violent criminals, it is routinely trumped by norms (within the law enforcement
community itself ) that encourage extrajudicial killing and protect those who engage
in it. Van Cott’s chapter on informal systems of justice administration in the Andes
offers additional examples. State law prohibits some acts (e.g., marital violence,
marriage of minors) that are permitted by indigenous law, and some transgressions in
indigenous law, such as gossip and religious dissent, are not crimes according to state
law. In these latter cases, community efforts to punish transgressors clearly infringe
upon individuals’ constitutionally enshrined rights.
Substitutive Informal Institutions. Finally, the upper right cell of figure I-1, which
combines ineffective formal institutions and compatible outcomes, corresponds to
substitutive informal institutions.∞∫ Like complementary institutions, substitutive
informal institutions are employed by actors who seek outcomes compatible with
formal rules and procedures. Like competing institutions, however, they exist in
environments where the formal rules are not routinely enforced. Hence, substitutive
informal institutions achieve what formal institutions were designed, but failed,
to achieve.
Substitutive institutions generally emerge where state structures are weak or
ineffective. For example, as Eisenstadt’s chapter shows, Mexico’s formal institutions
of electoral dispute resolution—such as the electoral courts—lacked credibility and
were frequently bypassed during that country’s protracted democratic transition. In
this context, concertacesiones, or ‘‘gentleman’s agreements,’’ thus served as a ‘‘way
station’’ for government and opposition elites until formal institutions of electoral
dispute resolution became credible. Similarly, in rural northern Peru, where state
weakness resulted in inadequate police protection and ineffective courts during the
late 1970s, citizens created informal rondas campesinas to defend their communities
and ronda assemblies (informal courts) to resolve local disputes (Starn 1999, 49–71,
106–32; Van Cott, this volume). These informal structures served a state-like func-
tion, dispensing community-level justice in areas in which the state had virtually
disappeared (Van Cott, this volume).
A few final points regarding this typology merit mention. First, whereas much of
the political science literature casts informal institutions as either entirely functional
introduction 17
negative effects of formal institutions. This dynamic can be seen in Chile (Siavelis,
this volume), where informal power-sharing arrangements helped attenuate the
effects of the Pinochet-era constitution, which in turn softened opposition to it
within the democratic Concertación. In such cases, the collapse of previously exist-
ing informal rules may trigger pressure for formal institutional change. For example,
Carey and Siavelis suggest in their chapter that the collapse of ‘‘electoral insurance’’
may generate pressure to modify Chile’s electoral system.
Finally, informal institutions may have a ‘‘crowding out’’ effect on formal insti-
tutions. Thus, substitutive informal institutions may inhibit the development of
effective formal institutions by dampening demands for the service provided by the
formal structure and encouraging actors to invest in (and thus gain a vested interest
in the preservation of ) the informal rules of the game.≤∑ For example, citizen invest-
ments of time, energy, and resources into informal justice systems in Andean com-
munities may contribute to the further neglect of (already weak) state-backed legal
systems.≤∏
This typology provides a useful starting point for categorizing the interaction
between formal and informal institutions.≤π However, at least two alternative types of
distinctions among informal institutions deserve mention. One is based on the
origins of informal institutions relative to formal ones. Some informal institutions
emerge endogenously from formal institutional arrangements, in that their origins
are directly related to particular formal rules. As we elaborate in the next section,
actors create formal rules in an effort to subvert, mitigate the effects of, substitute for,
or enhance the efficiency of formal institutions. Examples include many legislative,
judicial, and bureaucratic norms. Other informal institutions develop independent
of formal institutional structures, in response to conditions that are largely unrelated
to (and, in many cases, pre-date) the formal institutional context. Many indigenous
or customary laws fall into this latter category (Galvan 2004; Van Cott, this volume).
A second distinction, also discussed below, revolves around how informal institu-
tions emerge.≤∫ Some informal institutions are created in a ‘‘top-down’’ fashion by a
small number of elites. Among the informal institutions discussed in this volume,
the dedazo, as described by Langston, and the legislative and cross-party power-
sharing norms described by Carey and Siavelis, Siavelis, and Mejía Acosta fall into
this category. Other informal institutions emerge in a decentralized, ‘‘bottom-up’’
manner that involves a much larger number of societal actors. Clientelism (Des-
posato, Taylor-Robinson), many indigenous laws (Van Cott), and norms of police
violence (Brinks) fit this pattern. Whereas the dynamics of top-down or elite-created
informal institutions are in many ways similar to those of formal institutions, bottom-
introduction 19
up, ‘‘societal’’ informal institutions are often seen as linked to broader societal values
or cultural patterns.
institutions are a second-best strategy, it remains to be explained why actors who lack
the capacity to change the formal rules are nevertheless able to establish and enforce
informal ones. And where actors share certain illicit goals, it must be explained
how they are able to agree on informal norms that enable them to circumvent the
formal rules.
Along these lines, several of the volume’s chapters explore the creation of infor-
mal institutions through the lens of coordination. Because coordination often takes
place in a context in which power and resources are unevenly distributed, informal
institutions are cast as the culmination of a bargaining process in which actors seek to
maximize their benefits, given their beliefs about the strategies available to other
actors (J. Knight 1992). This is the gist of the arguments in the chapters by Langston,
Eisenstadt, Mejía Acosta, and Samuels.
Alternatively, the emergence of informal institutions may be explained as a histor-
ically contingent, and ultimately path-dependent, process.≥∏ In these cases, informal
rules are less a product of actor design than the (often unintended) consequence of a
particular historical experience that creates certain socially shared expectations.≥π
For example, as Stokes suggests in her chapter, norms of democratic accountability
may emerge out of an early felicitous experience with good government, which sets
in motion a virtuous cycle in which citizens believe that politicians can be held
accountable and, because of this, politicians are willing to act more responsibly.
Explaining how informal rules emerge and persist also requires specifying how
they are communicated to the relevant actors. Two mechanisms of the transmission
and enforcement of informal rules emerge out of the chapters. One is personal
networks, often operating through organizations such as political parties. Thus,
political parties played an important role in diffusing rondas campesinas in Peru
(Starn 1999, 116–17; Van Cott, this volume) and concertacesiones in Mexico (Eisen-
stadt, this volume), and in consolidating electoral clientelism in Honduras (Taylor-
Robinson, this volume).
Informal rules may also be communicated through highly visible (if infrequent)
episodes of rule-breaking and sanction. Widely observed efforts to punish deviations
from informal rules can effectively signal the costs of noncompliance. For example,
as Langston’s chapter shows, the Mexican dedazo was institutionalized during the
1940s and 1950s through a process of ‘‘learning by example.’’ PRI leaders who defied
the incumbent president’s right to choose his successor suffered political marginali-
zation, while those who played by the rules were rewarded with better posts. Like-
wise, in his chapter Brinks suggests that the murder of investigators into (or witnesses
of ) police crime communicated the norms of police impunity, effectively discourag-
ing others from taking such actions.
22 gretchen helmke and steven levitsky
Informal institutions are often portrayed as highly resistant to change. Like cul-
ture, they often are assumed to possess a ‘‘tenacious survival ability’’ (North 1990, 45),
which allows them to endure regardless of the formal institutional context. When
change occurs, it is expected to be slow and incremental (North 1990, 45; Lauth
2000, 24–25). Lauth, for example, argues that because informal rules ‘‘do not possess
a center which directs and co-ordinates their actions,’’ informal institutional change
is bound to be an ‘‘extremely lengthy’’ process (2000, 24–25). The chapters in this
volume, however, suggest that informal institutional change—both in highly cen-
tralized and in decentralized instances—may be more frequent than we often as-
sume. As Langston’s and Eisenstadt’s chapters on Mexico make clear, informal
institutions can collapse quite quickly. And, as Van Cott argues in her essay, many
indigenous institutions that are widely assumed to be long established as unchang-
ing have in fact been repeatedly transformed—and even reinvented—over time.
One possible explanation for variation in the stability of informal institutions lies
in the type of institution being examined. For example, ‘‘top-down’’ or elite-created
informal institutions, which are usually a product of strategic interaction among a
relatively small number of actors, may be more susceptible to change than ‘‘bottom-
up,’’ society-wide informal institutions, which emerge in a more decentralized (and
less conscious) manner through repeated interaction.≥∫ Because coordination is
often harder to achieve in larger groups, once an informal norm is in place, reorient-
ing expectations around a new set of rules may prove more difficult. This may be
too neat a dichotomy, however. Under certain conditions, even deeply rooted and
decentralized societal norms may change relatively quickly (e.g., foot-binding in
China; see Mackie 1996). Moreover, community size can cut both ways. As J. Knight
(1992) points out, informal institutions may actually be less stable in larger commu-
nities, given higher relative costs of sanctioning and increased opportunities for the
development of multiple interpretations of the norm.
There is a clear need, then, for greater attention to the questions of why and how
informal rules change. Explorations into informal institutional change will almost
certainly force scholars to think more seriously about their foundation. Identifying
potential sources of change in informal institutions is, in effect, the flip side of specify-
ing the mechanisms of institutions’ reproduction or stability (cf. Thelen 1999, 399).≥Ω
This volume’s essays consider three sources of informal institutional change:
(1) formal institutional change, (2) changes in the underlying distribution of power
and resources, and (3) changes in shared beliefs and collective experiences.
introduction 23
volume), and the increased effectiveness of Peru’s public security and judicial sys-
tems led to the eventual collapse of many rondas campesinas (Starn 1999, 265–68;
Van Cott, this volume).
The destruction of informal institutions does not always result in a stable new
formal institutional equilibrium, however. It may also lead to chaos. In Guatemala,
for example, state attempts to impose national legal institutions in rural Mayan
communities during the 1970s and 1980s disrupted preexisting patterns of customary
law, but the failure to consolidate a new legal system left rural residents uncertain
over which laws and authorities prevailed (Handy 2004, 555–60). As a result, ‘‘much
of Guatemala was essentially left lawless,’’ leading to a sharp increase in violence and
vigilantism (Handy 2004, 558–60; see also Galvan 2004).
Change in the Distribution of Power and Resources. Formal institutions are not
the only impetus for informal institutional change. An alternative source of change
is an alteration of the status quo conditions that had previously sustained the institu-
tion (J. Knight 1992; see also Greif and Laitin 2004). Developments in the external
environment may change the distribution of power and resources among actors,
weakening those who benefit from a particular informal institution and strengthen-
ing those who seek to change it. This dynamic can be seen in Langston’s analysis of
the collapse of the dedazo. Mexico’s increasingly competitive electoral environment
during the 1990s strengthened local PRI leaders and activists vis-à-vis the national
leadership, which allowed them to contest and eventually dismantle the dedazo
system. In this case, the relative pay-offs to PRI politicians were transformed by the
leadership’s growing inability to credibly threaten those members who challenged
the president’s prerogatives. Similarly, Eisenstadt argues that the shift in the balance
of power created by the PAN’s capture of the presidency in 2000 put an abrupt end to
concertacesiones. Informal postelection bargaining tables had always been a second-
best strategy for PAN leaders, and when the power asymmetries created by the PRI’s
control of the presidency disappeared, the PAN was free to abandon them. Finally,
Carey and Siavelis argue that the viability of the Concertación’s ‘‘electoral insur-
ance’’ hinges on the continued electoral success of the coalition. Were the Concerta-
ción to lose the presidency, it would lose its capacity to reward unsuccessful legisla-
tive candidates.
tors to sponsor locally oriented pork-barrel projects. She, too, finds statistical evi-
dence to support this hypothesis.
A second strategy is to focus directly on actors’ expectations about the informal
rules of the game. By examining actors’ mutual understanding of the rules, one can
distinguish more carefully between informal behavior patterns and informal institu-
tions. Ethnography is an important research tool in this task.∂≥ Several analyses in
this volume (including those by Brinks, Desposato, Freidenberg and Levitsky, Mejía
Acosta, Samuels, Siavelis, and Van Cott) draw heavily on ethnographic research,
identifying shared expectations about informal rules through extensive interviews
with (and observation of ) the actors who are affected by those rules. Such case
expertise is invaluable for understanding how the actors themselves understand the
informal constraints they face. Yet other methods may also be employed to get at
actors’ expectations. Stokes’s chapter, for example, offers a novel use of survey re-
search techniques to establish shared expectations. Rather than merely gauging
citizens’ attitudes and values, she designs survey questions to investigate whether
citizens in different localities in Argentina hold different expectations about whether
fellow voters will punish politicians who behave dishonestly.
A third approach to identifying informal institutions is to focus on mechanisms of
enforcement. If informal behavior is rule-bound, then violations of the rule must
trigger some kind of external sanction. Unlike formal sanctioning mechanisms (i.e.,
legal systems), informal sanctioning mechanisms are often subtle, hidden, and even
illegal. They range from different forms of social disapproval (hostile remarks, gossip,
ostracism), to the loss of employment, to the use of hired thugs and other means of
extrajudicial violence.
As Brinks and Samuels note in their chapters, a problem with identifying infor-
mal institutions through incidents of enforcement is that when they are functioning
well, enforcement is rarely necessary.∂∂ Still, even rare instances of deviation and
punishment can be telling. For example, Samuels shows how Brazilian politicians
who fail to deliver government contracts to their financial supporters have difficulty
raising money in future elections, and how politicians are able to delay or block
government contracts to firms that provide insufficient financial support. He shows
how the Collor government blacklisted entrepreneurs who had previously financed
Collor’s presidential bid but balked at another round of donations after his inaugura-
tion. Later, when Collor broke the rules and began to investigate corruption outside
his inner circle, many of the same politicians and entrepreneurs responded by
supporting his impeachment. Similarly, Mejía Acosta demonstrates how ‘‘going
public’’ about illicit vote-buying activities serves as a mechanism of enforcing ghost
coalitions in Ecuador’s legislature. He shows that when the Durán Ballen govern-
introduction 27
ment deemed that the Social Christian Party’s (PSC) demands had escalated beyond
the terms of the original ‘‘ghost coalition’’ pact, Vice President Alberto Dahik pub-
licly accused PSC leaders of corruption. In response, the PSC launched an im-
peachment drive that forced Dahik into exile. Likewise, Langston shows how PRI
executives used their control over state resources and electoral institutions to ensure
that defectors would lose elections and pay an enormous cost in terms of their
political careers. After defectors failed on three successive occasions during the 1940s
and 1950s, the threat of sanction became sufficiently credible that no major PRI
politician broke the rules for nearly three decades.
In a slightly different vein, Brinks’s chapter focuses on ‘‘permissive’’ rules, or
informal institutions that allow, but do not require, certain behavior. Under these
rules, sanctioning operates in a somewhat different manner. Actors may avoid the
permitted behavior without fear of sanction. Instead, punishment is meted out to
those who seek to enforce the formal rules that prohibit the behavior in question
(i.e., ‘‘whistleblowers’’). Thus, Brinks finds that judges, prosecutors, and police inves-
tigators who seek to enforce laws prohibiting extrajudicial killing by police face
harassment by superiors, noncooperation by the police, and even death threats.
The above discussion suggests that efforts to identify and measure informal in-
stitutions require substantial knowledge of the communities within which those
rules are embedded. Although this is certainly true, it would be a mistake to identify
informal institutional analysis only with case study research. In fact, the chapters in
this volume engage in a variety of innovative methods. For example, many of our
contributors use subnational comparisons to increase the number of observations or
cases while holding a variety of national-level variables constant.∂∑ Thus, Desposato
pairs two Brazilian states with similar formal legislative and electoral institutions but
different levels of clientelism, thereby setting up a ‘‘natural experiment’’ that allows
him to isolate the independent effects of clientelism on legislative behavior.∂∏ Brinks
engages in both subnational and controlled cross-national comparisons. His study of
five urban areas in Argentina, Brazil, and Uruguay produced evidence of informal
norms of police violence in Salvador (Brazil), São Paulo (Brazil), and Buenos Aires
(Argentina) but not in Córdoba (Argentina) and Montevideo (Uruguay). Stokes uses
subnational-level survey data to compare localities in which informal institutions are
presumably strong with those in which they are not.
Other chapters increase the number of observations by comparing single cases
across time. For example, Langston examines evolution of the dedazo across several
presidencies, from its establishment under Lázaro Cárdenas in the 1930s to its col-
lapse under Ernesto Zedillo in the 1990s. Through this comparison, Langston is able
to show how increased electoral competition helped erode the foundations of the
28 gretchen helmke and steven levitsky
dedazo. Similarly, Eisenstadt examines the use of concertacesiones across four time
periods, which allows him to show how changing political conditions (particularly
the increased credibility of formal electoral institutions and the rise to power of the
PAN) led to the demise of informal bargaining tables.
Finally, as the chapters by Carey and Siavelis, Stokes, and Taylor-Robinson dem-
onstrate, large-n statistical analysis may be employed both to identify informal in-
stitutions and to test propositions about their effects.
This introduction has sought to provide an initial framework for studying how infor-
mal institutions shape democratic politics. The chapters that follow build on this
framework. Drawing on diverse methodological tools, the contributors explore how in-
formal rules shape and constrain political actors, how they interact with formal demo-
cratic institutions, and what they mean for the quality and stability of democracy.
Part I focuses on executive-legislative relations. Chapter 1, by Siavelis, examines
informal power-sharing norms in post-Pinochet Chile. Chile’s formal democratic
institutions are among the least conducive in Latin America to effective governance.
Yet Chile is widely viewed as one of the most successful third-wave democracies.
Siavelis contends that three sets of accommodating informal institutions contributed
to democratic governability by facilitating cooperation within the governing Con-
certación: the cuoteo (a formula to distribute candidacies and government posts), the
partido transversal (a supra-party core of elites within the governing coalition), and a
pattern of informal settlements known as democracia de los acuerdos (democracy by
informal agreement). Chapter 2, by Desposato, examines the impact of clientelism
on legislative organization and behavior. Comparing two Brazilian states with vir-
tually identical formal institutions but significant differences in the level of clientel-
ism, he shows how the presence of clientelism can dramatically shape how legisla-
tive parties and individual legislators operate, even reversing the incentives provided
by formal institutions. Chapter 3, by Mejía Acosta, examines how legislative ‘‘ghost
coalitions’’ facilitated democratic governance and economic reform in Ecuador. He
argues that even though severe party fragmentation made Ecuador a ‘‘least likely’’
case for successful democratic reform, ghost coalitions—an informal institution in
which opposition parties exchanged support for patronage and pork but did not
publicly align with the government—enabled presidents to pass major legislation
while protecting legislative allies’ reputations from charges of collaborating with an
unpopular government.
Part II focuses on informal institutions and electoral politics. Chapter 4, by Sam-
introduction 29
uels, examines informal campaign finance contracts in Brazil. He argues that the
familiar elements of the logic of ‘‘credible commitments’’—reputation, iteration, and
the possibility of punishment—can support informal campaign finance contracts,
but that these conditions are most likely to be sustained within particularistic rela-
tionships, such as personal and family networks. Chapter 5, by Taylor-Robinson,
explores how incentives created by clientelism affect the behavior of elected officials
in Honduras. She argues that entrenched norms of clientelism may induce politi-
cians to behave in ways that are at odds with their interests as defined by formal
electoral institutions. Using data about the propensity of Honduran legislators to ini-
tiate locally targeted bills, Taylor-Robinson finds that clientelism creates incentives
for legislators to initiate local pork-barrel legislation even where electoral institutions
create no incentive to attend to local needs. Chapter 6, by Stokes, examines informal
norms of electoral accountability in Argentina. Although accountability in national-
level Argentine politics is usually seen as low, Stokes finds substantial within-country
variation. In some cities, such as Mar del Plata, shared expectations around a rule on
retrospective voting decisions—whereby voters weigh politicians’ past performance
more heavily than personalities or particularistic favors—have enhanced both ac-
countability and the overall performance of democratic institutions. Stokes then
offers some thoughts about the origins of such norms.
Part III focuses on party politics. Chapter 7, by Langston, examines the origins,
institutionalization, and eventual collapse of the Mexican dedazo, in which the
exiting president personally chose his successor for the top executive office. She
shows how sitting PRI executives devised rules to deter ambitious PRI politicians
from independently seeking the presidency, and how they used the resources af-
forded by hegemonic politics to impose steep costs on those who broke the rules. As
electoral competition grew in the 1980s and 1990s, however, the ability of PRI
presidents to enforce the dedazo eroded and the institution collapsed. Chapter 8, by
Carey and Siavelis, looks at ‘‘electoral insurance’’ in Chile. Due to the country’s
M=2 electoral system, Chilean parties and coalitions must put their strongest candi-
dates in the most precarious electoral list positions in order to secure legislative
majorities, which generates a divergence of interests between coalitions and politi-
cians. During the 1990s, the governing Concertación resolved this dilemma through
the informal practice of providing appointed posts for candidates who accept per-
sonal risk on the coalition’s behalf and run good—but just not good enough—
campaigns for congress. Although this insurance system helped hold the coalition
together for more than a decade, changes in the post-2000 electoral environment
may threaten the Concertación’s control over the appointed posts required to ‘‘pay
for’’ the informal institution, which may hamper its prospects for survival. Chapter 9,
30 gretchen helmke and steven levitsky
peter siavelis
Chile’s formal political institutions are regularly categorized as among the least
conducive to effective democracy in Latin America. The country has an exaggerated
presidential system, a weak legislature, a majoritarian electoral system, a partially
appointed senate, and high quorums for constitutional reform. In addition, despite
the efforts of military reformers, Chile remains a multiparty system. The compara-
tive theoretical literature suggests that this undesirable institutional combination
should create disincentives for cooperation, coalition formation, and political ac-
commodation. Presidents should have trouble legislating, particularly given Chile’s
multiparty system, and we should expect presidents to strong-arm or step outside the
legislature to advance their agendas. Theory also suggests that such an institutional
framework should ultimately lead, at best, to some type of delegative democracy,
and at worst to deadlock and possible democratic breakdown (Mainwaring 1993;
Linz 1994; O’Donnell 1994; A. Valenzuela 1994; Jones 1995; Mainwaring and Shu-
gart 1997).
Nonetheless, Chile’s democratic transition is recognized by scholars as among
the most successful in Latin America, and in regional perspective the country has
one of the best records of democratic governability. Following Chile’s return to
democracy in 1990, a stable pattern of two-coalition competition quickly emerged
between an alliance of the center-left (the Concertación) and an alliance of the right
(the Alianza por Chile). The Concertación alliance, which has governed since the
return of democracy, initially was cobbled together from more than a dozen parties
and has been the most stable coalition in the country’s history. Presidents have
34 peter siavelis
successfully legislated despite lacking majorities in congress and have not wielded
the wide-ranging powers granted them by the constitution. Rather, they have worked
with congress and the opposition to constructively negotiate their legislative pro-
grams (Siavelis 2002b). Instead of the deadlock and difficulty scholars would predict,
Chile embodies the best of representative democracy and consensus politics in Latin
America. Formal institutional analysis sheds little light on the reasons for this suc-
cess, making for quite a puzzle in explaining Chile’s democratic success from this
perspective.
Part of the solution to the puzzle lies in understanding informal institutions. A
complex network of informal institutions helped attenuate the problems that might
otherwise have developed as a result of Chile’s awkward institutional arrangements,
and also helped moderate the actions of presidents who have the formal powers to be
quite authoritarian. In particular, Chile’s democratic performance has been en-
hanced by what Helmke and Levitsky (in their introduction to this volume) call
accommodating informal institutions. Elites had strong incentives to create accom-
modating informal institutions to allow them to achieve goals within an institutional
framework that creates an uncomfortable fit and militates against many of their
fundamental interests.
In theoretical terms, evidence from Chile supports Helmke and Levitsky’s con-
tention that accommodating informal institutions are most likely to emerge where
strong effective formal institutions exist. However, it also underscores additional
conditions that facilitate the emergence of accommodating informal institutions.
Informal institutions are most likely to be found where political actors face difficulty
operating within formal institutions, or where there is a lack of congruence be-
tween political reality and formal institutional arrangements. Informal institutions
are more likely where all actors gain equally from their creation, where there are
shared expectations about potentially negative and positive outcomes, and where the
shadow of the future makes their maintenance worthwhile for the long term. Finally,
the notion of ‘‘informality’’ has a long trajectory in the literature on Latin American
politics and is most often expressed in terms of the negative consequences of nepo-
tism, patron-client relations, corporatism, and patrimonialism (e.g., Hagopian 1993;
Hillman 1994; Wiarda and Kline 1996). Less critical and normative treatments of the
importance of informal institutions are not as prevalent. This dim view of informal
institutions finds its roots in a general tendency to view politics in the developing
world as somehow dysfunctional if it does not conform to the norms of political
processes in developed countries. Evidence from Chile shows the very positive role
that informal institutions can play.
informal institutions and chilean democracy 35
legislate and govern, the ability to govern relied on coalition maintenance, and
coalition maintenance was contingent upon the capacity of political elites to con-
struct mechanisms for the widespread and fair representation of parties in govern-
ment and policymaking. The formal institutional structure militates against all of
these goals. In addition, the very high constitutional thresholds for changing the
system eliminated reform as an option for elites to make the institutional structure
better fit their needs. How did Chilean elites solve this dilemma? In essence, this
chapter argues that informal institutions made this ‘‘difficult combination’’ much
less difficult.
CHAMBER OF DEPUTIES
PDC 38 31.7 37 30.8 38 31.7 23 19.2
PPD 16 13.3 15 12.5 16 13.3 20 16.7
PS — — 15 12.5 11 9.2 10 8.3
PRSD 5 4.2 2 1.7 4 3.3 6 5.0
Others 10 8.3 1 0.8 0 0.0 3 2.5
Concertación total 69 57.5 70 58.3 69 57.5 62 51.7
RN 29 24.2 29 24.2 23 19.2 18 15.0
UDI 11 9.2 15 12.5 17 14.2 31 25.8
Others 8 6.7 6 5.0 6 5.0 8 6.7
Alianza total 48 40.0 50 1.7 46 38.3 57 47.5
Others 3 2.5 — — 5 4.2 1 0.8
Total 120 100 120 100 120 100 120 100
SENATEa
PDC 13 27.7 14 29.8 14 29.2 12 24.5
PPD 1 2.1 2 4.2 4 8.3 3 6.1
PS — — 4 8.5 2 4.2 5 10.2
PRSD 3 6.4 1 2.1 — — — —
Others 5 10.6 — — — — — —
Appointedb — — — — — — 4 8.2
Concertación total 22 46.8 21 44.7 20 41.7 24 49.0
RN 13 27.7 11 23.4 7 14.6 7 14.3
UDI 2 4.2 3 6.4 5 10.4 9 18.4
Others 1 2.1 3 6.4 6 12.5 2 4.1
Appointedb 9 19.1 9 19.1 10 20.8 7 14.3
Alianza total 25 53.2 26 55.3 28 58.3 25 51
Others — — — — — — — —
Total 47 100 47 100 48 100 49 100
Source: Author’s calculations with data from www.elecciones.gov.cl.
Notes: PDC, Christian Democratic Party (Partido Demócrata Cristiano); PPD, Party for Democracy (Partido
Por La Democracia); PRSD, Radical Social Democratic Party (Partido Radical Socialdemócrata); PS, Socialist
Party (Partido Socialista); RN, National Renewal (Renovación Nacional); UDI, Independent Democratic Union
(Unión Demócrata Independiente).
a All elective senate seats were filled in the 1989 elections. However, only one-half of the senate is elected every
four years. For 1993, 1997, and 2001, data reflect final membership of the senate.
b While appointed senators have no obligation to support the right, they are listed with this sector until 2001
because they were appointed from the ranks of the military or institutions influenced by it. Voting records show
they sided with the right, providing effective veto power on especially controversial legislation. Beginning in 2001,
some of the designated senators were appointed from the ranks of the governing parties and are listed with them.
Two former presidents with constitutionally mandated ‘‘life seats,’’ are counted along with their respective
ideological sectors (Concertación President Eduardo Frei and former General Augusto Pinochet—though he
does not actively serve in the senate).
informal institutions and chilean democracy 39
vided presidents with the tools to solve these twin problems. What is more, each of
the institutions built upon and relied upon the others. For example, democracia de
los acuerdos would have been impossible without the partido transversal, which
would in turn have been impossible without the cuoteo.
One might contend that there is a functionalist tone to the arguments set out
here, which would suggest that informal institutions simply ‘‘emerged’’ because they
were ‘‘necessary.’’ Coalition maintenance was, obviously, a collective good. How-
ever, the individual good of parties, it would seem, should at times trump this
collective good, creating incentives for parties to go it alone as collective-action
problems emerge. Nonetheless, this type of stark, rational choice analysis of the
problem overlooks the complexity of the multilayered game of postauthoritarian
Chilean politics, which provided sufficient incentives for actors to purposefully and
simultaneously act to resolve multiple collective-action problems. Informal institu-
tions did not ‘‘emerge’’ in a functional manner; they were built by political elites with
the goals of coalition maintenance and effective government in mind.
First, regime maintenance was a systemic collective good with very costly collec-
tive and individual sanctions. Going it alone and noncooperation were an extraor-
dinarily high-risk strategy, given that players very clearly knew the consequences of
interparty wrangling, coalition dissolution, and the governing incoherence and in-
stability they would produce. Regime crisis would mean the end of the democratic
game for everyone. Though in hindsight this may seem absurd, at the time it was a
very real possibility. The military was put on a state of alert at several key moments in
the transition, and on very high alert as late as 1993.
Second, one could contend that elites faced a similar threat of regime dissolution
in all of Latin America’s third-wave transitions, yet failed to resolve collective-action
problems as successfully as in Chile. However, echoing Helmke and Levitsky’s
contention about the importance of effective formal institutions in facilitating the
functioning of informal ones, Chile’s strong formal institutions provided a more
fertile ground for the growth of informal institutions than did those that existed
in other countries. Also, the unusual (in Latin American terms) structural char-
acteristics of Chilean groups and social organizations further facilitated the build-
ing of accommodating informal institutions. Weyland shows that, in general, Chil-
ean parties, unions, business associations, and other ‘‘encompassing organizations’’
are significantly more cohesive than those in neighboring countries, giving them
more capacity to punish free riders and effectively solve collective-action problems
(1997, 39).
Third, postauthoritarian cooperation was an iterated game in which repeated
interactions among elites built trust and underscored the benefits of coalition mainte-
40 peter siavelis
nance. Parties received individual benefits from participation in the collective game,
and these benefits underwrote continued participation. The tremendous success of
Concertación governments further contributed to this self-reinforcing dynamic.
Finally, the postauthoritarian Chilean political system was not a zero-sum game,
in which gains for parties represented proportional losses for the coalition. Large and
small parties had different goals, which amounted to a win-win situation that en-
hanced the incentives for coalition formation and maintenance and, hence, for the
purposeful creation of informal institutions to achieve these goals. Large parties
needed small parties on board to maintain a semblance of unity and ensure unified
presidential candidacies, and small parties simply wanted to participate in the gov-
ernmental game. In this sense, the informal institutions to which this analysis now
turns were successful primarily because they reconciled the collective and individ-
ual goals of politicians, parties, and the governing coalition.
the cuoteo
the simple splitting of candidacies. There is a consistent set of informal rules and
norms that govern candidate placement and respond to the constraints of the formal
electoral system. It would seem quite easy to simply divide candidacies between the
Concertación’s two subpacts, which have been relatively equal in levels of support
since the return of democracy. Indeed, the informal cuoteo begins with the assump-
tion that candidacies are shared equally between the subpacts of the left and the
center, and negotiation and tinkering proceeds from this assumption of parity.
However, the shared understandings of this informal institution do not end there.
Minor parties are crucial for coalitions to rally enough support to pass thresholds
nationally and to ensure the ability to maintain single coalitional presidential can-
didacies. But small parties cannot win by running alone. Further, as Carey and
Siavelis note (this volume), it is crucial for candidates not only to be placed on lists,
but to be placed on lists with a partner they can potentially beat. With district
magnitudes of two, a minor-party candidate paired with a coalition colleague from a
major party stands to lose, as each coalition is likely to win only one of the two seats
in the district. Small parties want not just candidacies, but seats they can win. In fact,
summing up the logic that prevails for small parties paired with larger ones, one
small-party negotiator who was locked in a struggle with a larger-party partner re-
plied to a very generous offer of candidate slates (in all of which the smaller party was
likely to lose!), ‘‘Don’t give me slates, I am asking for seats.’’π Therefore, the cuoteo
also provides that small-party candidates be paired with other small-party candidates,
or weak major-party candidates, to allow for some victories for minor parties (and
probably an overall lower district vote total than two major-party candidates would
garner). This is a counterintuitive logic from the perspective of coalition victory. It
means that the candidate capable of garnering the highest number of votes in a
district may be replaced by a weaker candidate in the interests of coalition unity.∫
How do major parties distribute this cost of ensuring unity? After beginning with
an assumption of evenly dividing candidacies between subpacts, parties understand
that the share of the seats they must cede to smaller-party partners depends on how
well they have performed in previous elections and where they stand according to
public opinion data. If a major party has a drop in support, it is understood that this
party will be the one within its subpact to cede seats to smaller partners, assuming the
costs of including small parties in the coalition.
These dynamics are well understood by adherents to the informal institution and,
in turn, lead to the type of enforcement mechanisms that are typically employed.
The main enforcement mechanism is withdrawal or threatened withdrawal from the
coalition or, for small parties, switching subpacts. Throughout the 1990s the Union
of the Center-Center (UCC) flirted with both coalitions in an attempt to extract seats
44 peter siavelis
and, indeed, in 2001 the Radical Social Democratic Party (PRSD) switched from the
center-left subpact to the left subpact in order to cut a better deal. Debates play out
both in closed meetings and in the press, where parties contend that they will
definitively cede no more seats, or that if they are pressured to surrender additional
seats it will mean the end of the coalition.Ω Still, each of the players needs the other,
and the consequences of failure to reach an agreement are high and well understood
by negotiators.
Enforcement is also more complex, however, because it takes place on more than
one electoral level. It is not a simple question of wrangling over the quota of candi-
dacies for each subpact and party. Short of coalition dissolution, parties also threaten
sanctions regarding the placement of candidates, withdrawing previous concessions
if they lose a particular candidacy to a competitor. They also engage in horse-trading
that involves multiple and multilevel sanctions. Indeed, leaders sanction doubters in
their own parties who question whether the leadership has reached optimal seat-
share agreements. In the final hours of negotiation before the 1993 elections, one
party leader slammed his fist on the table and threatened the resignation of the entire
party executive council if party loyalists failed to approve the pact that he admitted
was ‘‘not the best,’’ and indeed might be ‘‘bad,’’ but was better than ‘‘committing
suicide’’ (La Segunda 1993). The use of the word suicide is common in the lead-up to
every election, as parties play chicken to see who will give in first. Indeed, one analyst
sums up both the difficulty of negotiations and the stakes for parties in not reach-
ing an agreement, arguing that the binomial system is a ‘‘time bomb treacher-
ously placed under negotiating tables of the two alliances every four years’’ (Otano
2001). The statement also suggests the centrality of informal institutions in defusing
this bomb.
Thus, in response to the inflexibility and lack of representativeness of the formal
electoral system, elites adopted commonly understood and enforceable informal
rules for the distribution of seats among the Concertación’s parties. The enforce-
ment mechanisms are internal and external. Internally, the ‘‘election insurance’’
solution (Carey and Siavelis, this volume) provides rewards for risk takers, while
the externally generated threat of coalition dissolution provides incentives to reach
agreements. By losing coalition membership, individual parties surrender both their
ability to compete successfully in elections and their access to executive branch
appointments.
The formal electoral system provides a concrete logic related to the relative party
strength and the placement of candidates that is well understood, and to which
parties have adapted through the use of informal rules of candidate placement.
informal institutions and chilean democracy 45
These rules have, in turn, shaped and affected the nature of the elaborate informal
strategies designed to simultaneously balance the goals of promoting party interests,
ensuring coalition survival, and winning political office. Rather than simply merg-
ing as the military expected, political parties have used candidate selection processes
as just one of many sets of informal institutions constructed to ensure their survival.
Elites devised informal institutions to deal with the complexity of successfully recon-
ciling the goals of candidates, individual parties, and the coalition.
The partido transversal is another informal institution that has been critical to
governing and the success of the Concertación coalition. The term partido transver-
sal refers to a shifting but consistent informal group of leaders, with crucial roles in
the first democratic governments, who define themselves more as ‘‘leaders of the
Concertación’’ than as leaders of their respective parties. Members of the partido
transversal are centered around presidents and hold key jobs in the most important
so-called political ministries. Though there is no formal organization or formalized
meeting of the partido, the actors themselves know who they are, and they structure
informal relationships among themselves, between their parties and the coalition,
and, as discussed later, with social actors whose input is crucial to democracia de los
acuerdos. Ignacio Walker, who served in the Ministry of the General Presidency
(SEGPRES) under the first postauthoritarian president, Patricio Aylwin, notes that
the partido’s members ‘‘correspond to informal networks that have . . . exercised a
strong influence under the three administrations of the Concertación, both in terms
of strategic design and the set of public policies that have been pursued’’ (2003, 5).
The partido transversal traces its origins to a troika of leaders who were Ayl-
win’s most important advisors: Edgardo Boeninger (Christian Democratic Party
[PDC]–SEGPRES), Enrique Correa (Socialist Party [PS]–Secretaria General de
Gobierno [SEGGOB]), and Enrique Krauss (PDC–Interior). The partido was origi-
nally centered in SEGPRES, an office Aylwin charged with structuring relation-
ships between the executive branch and congress, among the various ministries,
and between the government and social groups. As Boeninger’s earlier-cited memo
(1989) suggests, coordination among parties was recognized as crucial to the coali-
tion’s success and, in turn, the maintenance of democracy. Initial conversations
structured within SEGPRES evolved into more consistent communications outside
SEGPRES, among what Walker has termed a number of ‘‘complicit’’ like-minded
leaders from several parties engaged in a series of ongoing ‘‘conversations.’’∞≠ The
46 peter siavelis
Finally, the partido transversal allows the disparate parties of the Concertación to
negotiate as a unit (rather than as a squabbling multiparty committee) with other
social actors to ensure the passage of legislation and facilitating democracia de los
acuerdos.∞≥
The roots and the incentives surrounding the operation of the partido transversal
as an informal institution are similar to those of the cuoteo. The clear recognition
of the threat of an authoritarian regression, combined with the reality that re-
gime maintenance was a shared goal of the coalition and individual parties, pro-
vided sufficient incentive to solve what potentially could have been numerous
collective-action problems. However, while the threat of a military incursion may
have provided the impetus for the creation of the partido transversal, the continued
benefits it provides to both parties and the coalition are the incentives for its main-
tenance. Parties benefit from an influential voice in government and legislation,
while the coalition and the president are better able to pass legislation. Parties
gain effective agents to manage their relationship with the coalition, while the
coalition can rely on members to sell the coalition’s product to a sometimes skeptical
clientele.
These relationships shed light on the incentive structures that gave birth to and
maintain the partido transversal as a crucial accommodating informal institution.
For the core of the partido, rewards and appointments grow out of service to the
coalition, not to the party. While the president is certainly a member of a party, since
the return of democracy the president’s primary job has been to maintain the coali-
tion to achieve other goals. Therefore, presidents reward service to the coalition,
making members of the partido transversal especially valued and more likely to
enjoy the beneficence of presidents in the distribution of ministerial seats within the
cuoteo.
Chile’s awkward formal democratic institutions also combined to provide incen-
tives for the creation and maintenance of the partido transversal and condition some
of the external sanctions for its violation. Former Aylwin spokesman Enrique Correa
noted that ‘‘as long as there is multipartism and presidentialism, the norm will have
to be supra-partism.’’∞∂ He suggests that members of the partido recognize the dam-
aging consequences of the model of coparty government, in which different parties
negotiate with each other to maximize their own individual utility and are less likely
to solve collective-action problems. This was the model of Chilean coalition govern-
ment in the past, and particularly during the chaotic years of Salvador Allende’s
Popular Unity government. The point of departure for members of the partido
transversal is a desire to control the damaging consequences of the emergence of
coparty government.
48 peter siavelis
the executive branch also have met and consistently negotiated with members of the
opposition in relation to legislation on tax reform, labor standards, and the mini-
mum wage.
That said, contrary to what one would expect given the extent of presidential
powers in Chile, all three postauthoritarian presidents have been circumspect in the
use of democracia de los acuerdos. This is intriguing in an age when delegative
democracy is the norm in countries with formally far weaker presidencies than
Chile’s. As noted, Chilean presidents have almost complete control over the budget-
ary process. The president presents the national budget every year, and if congress
fails to approve it, the president’s budget enters into force (Constitution of the
Republic of Chile, art. 64). However, scholars of budgetary politics have noted that
although deputies are barred from formal consultation on the budget, in reality
members have a good deal of informal input in its formulation stage (Baldez and
Carey 1996, 18). Their influence continues during negotiations in legislative con-
ference committees of the two houses and in negotiations outside formal congressio-
nal institutions. Presidents routinely consult with legislators of their own parties and
members of the coalition to reinforce coalition unity and ensure that the budget is
acceptable to coalition partners.
More importantly, informal negotiation has also been central to securing the
votes of the opposition to pass budgets, even though presidents do not necessarily
need these votes. During negotiations on most budgets, the Concertación’s various
ministers of finance have met and negotiated with opposition leaders. Formally,
presidents have the capacity to simply implement their budget, but it makes little
sense to do so given the consequences this would have for the long-term capacity of
presidents to legislate. Because budget negotiations are only one piece of presidents’
legislative agendas, in order to continue to be able to legislate they must avoid
completely alienating the opposition (Siavelis 2002a). Hence, despite their impres-
sive formal budgetary powers, presidents consistently choose informal avenues of
negotiation and consultation to come to budget agreements that are acceptable both
to members of the presidents’ coalition and to the opposition. This is a stark depar-
ture from the familiar model of ‘‘delegative democracy’’ and the image of Latin
American legislatures as compliant rubber stamps.
Chile’s three postauthoritarian executives also have engaged in a series of regular,
yet informal, consultations with the business community, producer groups, and
members of the conservative media.∞∑ Silva is correct in noting that capitalists were
granted privileged access to the executive branch in order to stem the fears of
business elites that the neoliberal model would be abandoned; he goes as far as to
characterize this influence as ‘‘institutional veto power’’ (1992, 103). He notes that
50 peter siavelis
business elites lobbied directly with the executive branch on crucial legislation that
affected business, and that Chile’s main business lobby, the CPC (Confederación de
Producción y Comercio), was particularly influential in derailing initiatives that
were perceived to be antibusiness. While Silva characterizes this relationship as one
that sacrificed potentially progressive economic reform in exchange for political
democratization, these relationships also granted executives the power to persuade
conservative forces in congress that business interests had been taken into account in
elaborating budgets and controversial economic legislation.
Democracia de los acuerdos, then, helped stave off potentially destabilizing con-
flicts both within the coalition and between the coalition and the opposition. By
ensuring that the fundamental interests of allied parties, the opposition, and key
social groups were not threatened, Concertación presidents have succeeded in gov-
erning and in passing legislation in what is, in formal institutional terms, a very
infertile environment for consensus politics.
While one could contend that this pattern of conflict resolution is simply a
creative legislative strategy for presidents, the roots of democracia de los acuerdos as
an informal institution run deeper. First, there is a shared understanding among the
legislative right, powerful social groups, and Concertación governments on the rules
of the game. Concertación governments have informally acknowledged that social
groups have veto power over important legislation, and the government has re-
sponded with informal strategies to preempt potentially destabilizing opposition.
Now, important social groups expect to be informally consulted on controversial
legislation, resulting in an informal rule that for such legislation, reaching a pre-
legislative acuerdo is an integral part of the Chilean informal political game, with
consequences for violation.
Second, the rules of the game and the sanctions are clear. Entrepreneurs, the
conservative media, and active and retired members of the armed forces have pres-
sured the parties of the right on controversial legislation, and these powerful social
groups have important influence on the right’s parliamentary contingent. The tran-
sitional political system was a high-stakes game, in which presidential failure has
serious consequences. If presidents wanted to legislate, they would have to marshal
the support of at least some of the right, and failure to do so would have serious
consequences for presidents’ legislative agendas and, ultimately, for democracy. At
several stages in the democratic transition, rightist parties and interest groups made
clear that they feared the entire Pinochet economic model was in danger and that
this was unacceptable to them (Boylan 1996). By systematically engaging in a game
of democracia de los acuerdos, presidents assured business elites and the parties of the
right that their fundamental interests were taken into account.
informal institutions and chilean democracy 51
Informal institutions, then, explain a good deal of the success of Chilean elites in
solving the problems of exaggerated presidentialism in a multiparty system. Contrary
to what one would expect, elites formed coalitions, presidents legislated, and the
scope of presidential power was circumscribed. Still, this fails to answer the deeper
question of why and how Chilean elites built informal institutions.
What does Chile tell us about the conditions under which accommodating
informal institutions are created? Helmke and Levitsky argue in their introduction
that these types of informal institutions are more likely to emerge where effective
formal institutions already exist. Empirical evidence presented here both confirms
this assertion and sheds some additional light on the determinants of the emergence
of accommodating informal institutions. These, in turn, suggest several broader
theoretical implications.
In many cases the creation of accommodating informal institutions is tied to the
inability of political actors to solve problems or effect change within the context of
formal institutions. It is certainly costly to create and maintain informal institutions.
However, it is often more costly to change existing formal ones. Thus, most informal
institutions emerge when the cost of changing formal institutions exceeds the cost of
creating informal ones. Within the bounds of these general contexts, there are
several reasons why formal institutions can become less than optimal mechanisms
for solving problems, and informal ones are created.
The Chilean case shows that elites create accommodating informal institutions
where there is a lack of congruence between political reality and formal institutional
arrangements—that is to say, where political contexts make it difficult for elites to
52 peter siavelis
across contexts is that of how Chilean elites were able to overcome problems of
collective action to formulate accommodating informal institutions, when this en-
deavor has clearly failed in other contexts. The Chilean case is illuminating and
provides important tentative theoretical propositions about how the need for infor-
mal institutions translates into their concrete creation.
First, the formation and maintenance of informal institutions in Chile was an
elite project. It was the highest-level party leaders, presidents, presidential candi-
dates, and legislators who created the informal institutions and communicated the
rules of the game. We know from other chapters in this volume that many informal
institutions are society-wide and characterized by diffuse enforcement mechanisms.
Nonetheless, evidence from Chile suggests that accommodating informal institu-
tions that temper already strong formal institutions may be more likely to develop
and endure when a small number of actors is charged with creating them and
communicating their rules and sanctions for violations. Many of the more diffuse
informal institutions emerge spontaneously. Accommodating institutions, on the
other hand, are more purposefully created to correct for failures in often complex
formal institutional frameworks and are less rooted in long-standing, routinized
forms of behavior. Elite-created informal institutions may also be successful for other
reasons. Weyland (1997) notes how a pattern of elite political learning contributed to
the success of the Chilean transition because both the left and the right had learned
from painful past experiences during and after the military regime, and understood
the consequences of failure in this high-stakes environment.
Second, elites are more likely to create informal institutions when there are
shared expectations about potentially negative and positive outcomes. In terms of
regime crisis and military intervention, Chilean elites understood clearly that a
political crisis produced by partisan wrangling would bring a quick military interven-
tion. Thus, proactive efforts to stem conflict were essential, and the best way to do so
was to form coalitions that would decrease uncertainty. Because the formal institu-
tional structure created disincentives for accommodation, elites sought to improve
the incentive structure to better underwrite coalition formation and maintenance.
We might even go as far as to say that a shared external threat can provide incentives
for the creation of accommodating informal institutions as protective mechanisms.
As time passed, party elites also understood the consequences of failing to maintain
the coalition and, in turn, failing to put together joint electoral lists. There was
widespread understanding of the new constraints created by the electoral system,
which provided a powerful mechanism of enforcement for maintaining informal
institutions. This information was constantly communicated not only by political
54 peter siavelis
elites, but also by think tanks and partisan study groups that consistently provided
electoral simulations and other evidence to demonstrate the catastrophic electoral
consequences of failing to maintain coalitions.
Third, the Chilean case also suggests, however, that simple elite recognition of
threat may not be enough to solve collective-action problems related to regime
maintenance. Evidence from Chile suggests that informal institution-building by
elites may better succeed where social organizations are highly cohesive and better
able to sanction free riders. These types of organizations may have a greater capacity
to solve collective-action problems in the negotiation of posttransition settlements.
Fourth, informal institutions are more likely to form when all actors stand to
benefit more or less equally from their creation, and where there are few veto players.
The informal institutions that helped sustain Chilean coalitions benefited presi-
dents, party leaders. and legislators alike. Presidents and presidential candidates
benefited from the knowledge that they would not face intracoalitional competition.
Large political parties benefited from the knowledge that the addition of small
parties would help them cross crucial electoral thresholds. Small parties benefited
from the knowledge that they would receive a certain number of electoral slates and
some degree of representation in the ministries. The potential for veto players to
undermine coalition unity was minimized. Indeed, as the Carey and Siavelis chapter
in this volume shows, a reward system even developed for those actors who faced
sacrifice on the altar of coalition unity, but who otherwise might have upset the
delicate balance of agreements that sustained the coalition.
Finally, evidence from the Chilean case suggests that the ‘‘shadow of the future’’
and iterated interactions can help encourage the formation and maintenance of
informal institutions (Axelrod 1984, 174). For example, the subpact of the left knew
and expected that it would be awarded a presidential candidacy following the admin-
istrations of two Christian Democrat presidents from the center-left subpact. This
encouraged the left subpact to continue to respect the rules of the coalition game,
with the expectation that it would benefit from continued cooperation. At a lower
level, sitting ministers are hesitant to upset the partisan balance within ministries, to
ensure that there are places for them in future governments whose presidents might
be from other parties in the coalition.
conclusion
Chile inherited an institutional structure that in formal terms is among the least
conducive to cooperation, consensus, and the legislative success of presidents, and
that seems to be a recipe for delegative democracy and exaggerated majoritarianism.
informal institutions and chilean democracy 55
And yet, surprisingly, Chile is arguably the most successful of the third-wave Latin
American democracies. The analysis here has shown that this outcome cannot be
explained without reference to a series of accommodating informal institutions that
allowed elites to successfully achieve their individual and collective goals within
what is an unquestionably awkward and inflexible formal institutional arrangement.
Power-sharing in the form of the cuoteo, widespread consultation by way of the
partido transversal, and the adoption of an informal yet regularized pattern of demo-
cracia de los acuerdos provided Chilean elites with the tools to solve the many
collective-action problems created by multiparty presidentialism and to meet the
challenge of coalition maintenance in a multiparty presidential system.
There is little doubt that informal institutions will change where shared expecta-
tions change, and where the benefits to be gained by informal institutions begin to
be distributed unevenly. Indeed, talk of coalition dissolution has increased in Chile.
We know from the Carey and Siavelis chapter that Chile’s informal reward structure
is likely to change should the president not be from the Concertación or should the
relative power of parties change radically. Similarly, as the threat of the military
recedes we would expect a diminished incentive to adhere to some of the informal
rules that govern the operation of informal institutions, because the collective threat
that was so important in spawning them will recede. Relations within the Concerta-
ción are already more raucous than at any time in the past. It is difficult to say
whether this is the result of the age of the coalition or the charges of corruption
within it, or is a reflection of the changing fortunes of the country’s parties. However,
while the military threat may have provided the initial impetus for the creation of
informal institutions, a constant threat is not necessary to maintain them. If informal
institutions continue to meet the other conditions for the maintenance of informal
institutions, as set out here, their pure workability and demonstrated success may
induce the increased formalization of the accommodating informal institutions that
have been so central to Chile’s democratic success.
chapter two
s c o t t w. d e s p o s a t o
Informal institutions pervade political systems and far exceed formal institutions
in their age, number, and scope. Evolving work has made a compelling case for their
impact on the behavior of political actors, complementing formal institutions and
shaping equilibria where formal institutions are absent. But how much importance
should we ascribe to informal institutions? Do they just ‘‘fill in the gaps’’ where
formal institutions are absent? Do they disappear in the presence of strong formal
institutions? Or can informal institutions trump formal rules in shaping actors’
incentives?
In this chapter, I explore these questions by examining the effects of informal
electoral institutions on legislative politics in the case of Brazil. Recent scholarship
on that country has emphasized the importance of formal institutions in shaping
legislative politics, but mostly ignores the role of informal institutions in politics.
Through a comparative analysis of state assemblies, I show how informal electoral
institutions have important effects on legislative behavior, countering and even
reversing the impact of formal electoral institutions. In particular, I show how the
nature of informal electoral institutions—whether clientelistic or programmatic—
shapes the nature and structure of legislative behavior across states.
Previous work on Brazilian legislative politics focuses on the role that two formal
institutions play in shaping the legislative arena: federalism and electoral rules. On
brazilian legislative arena 57
the first of these, scholars have argued that the decentralization of resources in Bra-
zil has fragmented national power. In particular, federalism gives state-level actors
substantial influence over the behavior of national legislators. Brazilian federalism
grants subnational actors control over resources that are valuable for advancing
political careers. In particular, state governors control ‘‘pork,’’ campaign resources,
and even state ministerial jobs, all of which are coveted by national legislators. State
parties control nominations for all state and national legislative positions. Conse-
quently, national legislators should be responsive to pressure from state-level actors
that control career-enhancing and -advancing resources. The implication is that
national political parties are fragmented by state interests, and presidents must com-
pete with these interests to advance a national policy agenda (Mainwaring 1997,
1999; Selcher 1998; Souza 1998; Ames 2001; Samuels 2002; Desposato 2004).
The second institution, open-list proportional representation, is said to promote
personalism and weaken political parties. Under electoral rules of this type, citizens
cast a single vote, either for an individual candidate or for a party list (in Brazil, most
vote for an individual). After election results are tallied, seats are distributed in two
steps: first, to parties, in proportion to the share of the votes received by all their
candidates; second, within the parties to the top vote-getters. The easiest way to
achieve office under these rules is almost always to be one of the best vote-getters in a
party, and the easiest way to be at the top in a party is to run a campaign that attacks
others within one’s own party. Within-party competition for votes should be fierce,
and this competition, the argument goes, weakens national parties’ cohesiveness
(Mainwaring 1991, 1999; Pinheiro 1998; Graeff 2000; Ames 2001).
Informal institutions are largely excluded from the theory and empirical analysis
of both federalism and electoral systems. But evidence from other democracies
suggests that these formal institutional rules are not nearly a complete explanation
for politicians’ behavior. For example, both Chile and Finland use similar electoral
systems, but neither has been characterized as an ‘‘antiparty system,’’ as has Brazil. At
least one explanation for the difference could lie in informal institutions; for exam-
ple, the nature of partisan cleavages. This is not sufficient evidence that informal
institutions matter, however. In each case, one might also attribute these differences
to differences in formal rules. For example, Brazil has a federal system of govern-
ment, but Chile’s is unitary; Finland uses a parliamentary form of government, but
Brazil uses presidentialism. So cross-country differences might be the result of infor-
mal institutions or of other institutional features. Because there are many institu-
tional differences across all countries, it is difficult to isolate whether informal in-
stitutions help explain cross-country differences.
Ultimately, any cross-country comparison of institutions makes inferences about
58 s c o t t w. d e s p o s a t o
the effects of informal institutions difficult. When faced with differences in out-
comes across two countries with very similar formal institutions, those skeptical of
the explanatory power of informal institutions may suggest that subtler formal in-
stitutional rules may explain the disparities. Two countries may share presidential-
ism and single-member districts, but their nomination procedures or internal legis-
lative rules may differ substantially. This is why a within-country comparison of
Brazilian states is so useful. The states share nearly identical formal institutional
rules, including the same electoral rules, terms of office, nomination procedures,
basic form of government, and separation of powers. Differing dramatically across
states, however, are the informal electoral institutions. In some states, clientelism is
an important part of electoral politics; in others, it has been replaced by largely
programmatic electoral markets.
The Brazilian state legislatures provide a nearly ideal laboratory for testing the
impact of informal institutions. The states vary greatly in their informal institutions,
but share virtually identical institutional environments, most of which were imposed
by the national constitution. In this project, I chose two Brazilian states that capture
the extreme regional diversity of that country: Piauí and São Paulo. The first, Piauí, is
extremely poor, is less-developed, and has very clientelistic politics. São Paulo is
wealthier, has a more diverse and vibrant economy, and is generally accepted as
having a much smaller private goods component to elections than Piauí. Therefore,
if only formal institutions shape legislative behavior, there should be no differences
in legislative politics across states. But if societal variables—specifically, the presence
or absence of the informal institution of clientelism—matter as well, we should find
obvious differences across states.
This topic is of importance for several reasons. First, it provides a strong test of the
relative effects of informal and formal institutions. Most work on legislative politics
focuses on formal institutions as explanatory variables, including electoral systems,
procedural rules, and constitutions. Implicitly, these studies assume that all electoral
markets are alike or that voters and their characteristics do not matter. Consequently,
legislative institutional rules should have the same effects in any electoral context.
Relaxing and testing these assumptions is an important next step for political sci-
ence. Do legislatures work the same way regardless of the presence or absence of
informal electoral institutions? If not, we should be explicit and systematic in adjust-
ing institutional theories to fit different realities.
In addition, the chapter tests the idea that informal institutional effects can spill
over into multiple political arenas. Previous work has focused on the impact of
informal institutions in insular settings. For example, political scientists study infor-
mal institutions within legislatures, or vote-buying within electoral markets, but
brazilian legislative arena 59
we have not studied their spillover into other and broader political arenas, policy-
making, or democratic consolidation. This chapter shows how the effects of informal
institutions are not limited to local contexts, but these institutions can also directly
affect other political arenas.
I use the term electoral markets to define the mechanisms and patterns whereby
voters and candidates exchange votes and policy promises. The market is created by
formal institutions that endow citizens with votes and election winners with power to
direct government. Two idealized extremes of electoral markets are clientelistic and
programmatic. In clientelistic systems, candidates cannot enforce the behavior of
‘‘bought’’ votes. In programmatic systems, voters cannot enforce the provision of
policy goods. In each case (and in those in between), informal institutions help
make these markets function.
At one extreme, electoral markets may be cleared by the exchange of cash or other
private goods for votes—effectively, vote-buying or clientelism. This exchange is
enforced by either informal norms or illegal campaign activities. Payments are typi-
cally made before the election and ballots are usually secret, so candidates cannot
enforce exchanges. Citizens may agree to vote for a candidate in exchange for a cash
payment, but a candidate cannot enforce delivery of the promised vote. However,
informal norms and mechanisms have evolved to overcome the otherwise missing
incentives for compliance. Creative candidates have found quasi-enforcement mech-
anisms. For example, candidates might buy votes with new shoes, delivering one shoe
before and the other after the election. In other cases, voters simply have developed
norms of moral obligation to vote-buyers. Observers of, and participants in, vote-
buying schemes have effectively described a sense of gratitude and obligation to the
vote-buyers. In any event, with the secret ballot, vote-buying would disappear without
informal institutions that effectively enforce market mechanisms. Voters could easily
take payments and not vote for candidates—certainly some proportion of voters do
this. Ultimately, such an outcome could be seen as inefficient for voters and politi-
cians. Poor, risk-averse voters may well prefer private goods rather than policy prom-
ises, and noncompliance would, over time, destroy the market for such exchanges.
Where clientelism predominates in elections, politicians’ career survival depends
on obtaining and delivering private goods. In some countries, political parties control
access to clientelistic goods. But in many presidential systems, including Brazil,
executives are the best source for valuable campaign goods. Presidents typically
dominate spending, as do governors at the state level. They make initial budget
60 s c o t t w. d e s p o s a t o
proposals, often have exclusive bill-initiation powers for fiscal legislation, and control
the disbursement of appropriations (Schneider 2001). Government funds are not the
only sources for vote-buying, as candidates may use their own resources or donors’
funds. But, especially in poor areas, state coffers are the easiest and most abundant
source of campaign financing, direct or indirect.∞
At the other extreme, electoral markets may be driven by the exchange of votes for
broad policy promises. Candidates offer policy positions and voters support those
candidates whose proposals are closest to voters’ own. In programmatic environ-
ments, voters commit to candidates first and have no enforcement mechanism to
guarantee that candidates will deliver their promised policies. Typical understand-
ings of representation rely on enforcement through repeated and frequent elections
—politicians who do not deliver will soon lose reelection bids. In programmatic
systems, however, evaluating the performance of individual candidates is very diffi-
cult, certainly much more difficult than in clientelistic environments. Politicians’
work cannot be measured simply in terms of whether the environment has been
noticeably improved, corruption eliminated, or the International Monetary Fund’s
model of development defeated, because everyone knows that a single legislator
rarely can do all these things in a single term. So measuring and evaluating effort
requires observing credible policy moves toward that ultimate policy goal. Such
effort has many facets, including roll-call votes, committee work, and legislative
proposals. By contrast, with vote-buying, voters know immediately whether or not
candidates have complied with their promises. But how does one evaluate efforts to
fight globalization or work to implement the death penalty?
Ultimately, the challenges that legislators face—delivering goods or representing
policy positions—create very different challenges for strategic politicians. Clientelis-
tic informal institutions will increase pressure for the short-term delivery of private or
local public goods. Consequently, where electoral markets work through the infor-
mal institution of clientelism, legislators will focus on obtaining state resources for
distribution to voters and donors. In the case of Brazil, with strong executives, this
means that legislators should trade legislative power for access to resources. Execu-
tives should have solid and sizeable majority coalitions in legislatures, opposition
parties should effectively collapse as members begin to support the governor’s posi-
tion, and executive-legislative conflict should be minimized—all the result of incen-
tives created by informal institutions.
Programmatic informal electoral institutions will increase incentives for policy
efforts and aggressive credit-claiming. Legislators should work hard to advance plat-
forms, through legislation, committee work, and floor debates, and also to advertise
their efforts and positions as credible to constituents. Increased incentives for policy
brazilian legislative arena 61
São Paulo and Piauí, as noted earlier, represent polar extremes in Brazilian
politics. Piauí is one of Brazil’s poorest and most rural regions. This northeastern
state ranks among the most backward in Brazil on many indicators. In contrast, São
Paulo is Brazil’s economic engine, the center of modern industrial and financial
activity.
Table 2.1 provides a basic demographic profile of the two states on three dimen-
sions: percentage of residents with less than one year of education; percentage
earning less than the minimum wage; and percentage with running water plumbed
into their homes. There are stark differences between the two states on each dimen-
sion. The great majority of Paulistas have more than one year of education (87%),
while a majority of Piauíenses have less (53%). A small percentage of Paulistas live in
poverty (12%), but nearly two-thirds of Piauíenses live on less than the minimum
wage (about US$100 in 1998). Finally, almost all Paulista homes have running water
(96%), but only a third (34%) of Piauíenses enjoy the same.
62 s c o t t w. d e s p o s a t o
table 2.1
Demographic Profiles of São Paulo and Piauí, Brazil
(percentage of residents in state), 1990 Census
For many reasons, clientelistic electoral markets tend to appear in poor, less-
developed environments, while programmatic systems are more common in more
developed areas. Piauí and São Paulo are not exceptions to these trends. In Piauí,
clientelism has been an important part of elections, while in São Paulo, electoral
markets are much more programmatic. Note that neither state represents the ide-
alized extreme described above, where elections are completely determined either
by vote-buying or by highly charged and informed ideological voters. There are
ideological voters in Piauí, and there are vote-buyers and -sellers in São Paulo. But
overall, the two states occupy very different positions on the scale from one extreme
to the other. That is, although Piauí and São Paulo may not be idealized extremes of
pure clientelistic electoral markets or pure ideological electoral markets, they clearly
fall at different ends of that spectrum.
A few examples illustrate the differences between the two states. In Piauí, all
deputies that I interviewed confirmed that vote-buying is common, and all described
electoral relations based on the provision of local public or private goods. One
interviewee told me that many deputies were directly purchasing votes: ‘‘he [the
impoverished voter] wants a guardian angel so that when he is hungry, when the rain
is weak, the boss gives food and medicine. The boss resolves his principal problem:
dying of hunger, or a child dying from disease.’’≥ He also noted that while ten or so of
the thirty state deputies would directly purchase votes with payments, all deputies
provided services to constituents—access to public health clinics, say, or a chance to
enroll their child in a public school. One deputy’s major accomplishment was
getting a road built, but he noted that this was not enough to get himself reelected.
He also had to provide transportation for voters, medical care, and other private
goods. That is, local public goods such as roads were not sufficient for election;
voters demanded private goods.
Other interviews in Piauí produced similar results. Deputies emphasized the
importance of providing goods to voters or groups of voters. A politician might
finance a graduation party, or pay for a musical group to perform for the community,
or pay for a medical exam, x-rays, or other kinds of health care. These sorts of
brazilian legislative arena 63
A comparison of legislative roll-call behavior across Piauí and São Paulo reveals
substantial differences in the two states (table 2.2). First, political parties in São Paulo
are, on average, much more cohesive than those in Piauí: overall cohesion scores of
.86 versus .69. Both states use open-list proportional representation electoral rules,
which are highly personalistic and should cause intraparty conflict and low party
cohesion scores. But São Paulo’s score is much higher than Piauí’s. This difference is
consistent with an informal institutional explanation of legislative behavior. São
Paulo’s electoral markets are more programmatic than Piauí’s, increasing pressure
for, and payoffs from, cohesive political parties.
Second, in addition to open-list proportional representation, both states have
powerful governors. Given the combination of personal vote-seeking and budget-
controlling governors, party organizations should be trumped by state governors.
Opposition parties should have difficulty maintaining cohesion, as their members
support governors’ proposals in exchange for access to the state budget. The evi-
dence, however, again shows differences across states. In Piauí, all is as expected.
The governor’s coalition members have very high discipline (cohesion score .94) due
to the alignment of partisan and gubernatorial preferences. But the opposition par-
ties have virtually collapsed, with cohesion levels below .50. As expected, alignment
with the governor is much more important than party consistency where clientelism
is common. But the pattern reverses in São Paulo—opposition parties are slightly
more cohesive than government parties. Again, the difference between Piauí and
São Paulo reflects the incentives provided by informal institutions. In São Paulo,
increasing emphasis on programmatic goods in elections means that legislators focus
more on delivering policy, not pork. This informal institutional difference means
that governors have much less influence over roll-call votes in São Paulo than
in Piauí.
Finally, just comparing the frequency of roll-call votes across states shows how
informal institutions can dramatically change legislators’ behavioral incentives. São
Paulo had a fair number of roll-call votes: 1,042 over the period studied. Piauí had
only 20 such votes. The difference can again be ascribed to informal electoral
institutions. In Piauí, there are no payoffs for taking public positions on policy issues
brazilian legislative arena 65
table 2.2
Legislative Behavior in São Paulo and Piauí, 1991–98: Roll-Call Cohesion
Cohesion score
We will continue just as firm as before, but we will be able to vote in favor of the
government’s proposals.’’∑ It turned out that deputies in his own party were pub-
licly siding with the governor, and that others had simply switched to one of the
governmental parties in exchange for ‘‘benefits.’’ The PFL leadership had tried to
strengthen their party’s opposition to Governor Santo—but failed as its members
sided with the governor.
I asked opposition deputies in Piauí why they did not use verifications to obstruct
the government’s agenda. Some deputies simply made reference to ‘‘negotiation’’
with the governor and the majority, with no additional comment. I interpreted these
comments as indicators of legislative-executive negotiations, trading roll-call votes
for particularistic state resources.
I also spoke with opposition deputies from parties with reputations for disciplined
and ideological behavior (the Worker’s Party [PT] for example). The few deputies in
Piauí from these parties claimed that they did not know they could request verifica-
tions. Some seemed to like the idea—one said it was a way to ‘‘give names to the
cattle,’’ and others claimed they would begin to use this strategy in the legislature.
This may generate roll-call votes for use by scholars in the future; or the deputies’
responses might simply have been dishonest. Either way, the lack of roll-call votes
suggests that public position-taking or moving forward with a legislative agenda have
had little to do with legislators’ career strategies in Piauí.
Politics in São Paulo were distinctly different. Political parties in São Paulo were
relatively cohesive. Observers told me that most deputies follow their party leader’s
orientation for roll-call votes. The differences between Piauí and São Paulo were re-
flected in the way São Paulo dealt with controversial legislation. I observed the oppo-
sition using quorum calls, roll-call votes, and its allotted discussion time to prolong
both the 1999 budget and the privatization of Comgas (Gas Company of São Paulo)
processes as long as possible, with many sessions extending well past midnight.
But just as the São Paulo electorate is diverse, so are that state’s deputies. As
discussed above, some have attentive sectoral constituencies, but others have more
private-goods-oriented voting bases. Abrúcio (1998) noted how Governor Covas was
elected with a minority in the assembly, but quickly and easily assembled a coalition
—from deputies whose electoral strategy required private and local public goods.
One of the governmental party’s chief analysts acknowledged the emotional debates
in the state legislature, but qualified this by noting that none of the governor’s key
policy proposals had been defeated—all passed eventually.
While cohesion was relatively high in São Paulo, there is no conclusive evidence
that it reflects party discipline. Cohesion indicates that deputies from a party are
voting together, but discipline means that such voting is encouraged by the party and
brazilian legislative arena 67
that significant costs are imposed on deputies who refuse to conform. Party leaders
from the ‘‘catch-all’’ PMDB and Brazilian Progressive Party (PPB) both complained
to me about the lack of party discipline.∏ Deputy Tonin (leader of the PMDB) noted
that he had formal powers to discipline members, including control over their com-
mittee memberships, but suggested that these powers were of little value. He did
recount party meetings to discuss polemical roll-calls, where he would expound the
need to ‘‘stay together on this one.’’ Members of several parties told me they could
vote against their party if they chose to. Only members of the PT mentioned that
party cohesion was mandatory and that there were costs to voting against the party.
Finally, other interviews suggested that roll-call votes had nothing to do with disci-
pline or ideological cohesion. One observer, when discussing the legislative battle
over Comgas privatization, simply said: ‘‘The price must be very high,’’ referring
to the cost of deputies’ bargaining for state resources in exchange for support for
that bill.
My interviews in São Paulo left the impression that two mechanisms were at
work there. First, there was a relatively disciplined opposition that did not need to
seek private goods from the governor. These deputies, like those in the PT, had
public goods constituencies and did not need to bargain with the executive. Second,
deputies in the government’s coalition included both ideological supporters of the
governor’s agenda and goods-seeking bargainers. Further, because fewer deputies
needed private goods, the governor’s market for votes on controversial legislation was
smaller. As a result, the price of votes may have been higher, and prolonged negotia-
tions were sometimes necessary. In contrast to Piauí, where the governor dominated
the legislature, the São Paulo executive, though clearly influential, had to engage in
substantial negotiation to advance his agenda.
Of course, clientelistic politics are still important in São Paulo. Not all deputies
used roll-call votes in campaigns or speeches, or faced criticism of their voting
records on the campaign trail. But the fact that roll-call votes are frequent in São
Paulo and not in Piauí, and that deputies in São Paulo did use roll-call votes as part of
their career strategy whereas deputies in Piauí did not, is evidence of the impact of
informal institutions.
conclusion
This chapter offers several contributions to the study of legislative politics and
informal institutions, both in Brazil and more broadly. First, by controlling for the
many confounding variables that can plague cross-country studies, it provides strong
evidence for the importance of informal institutions. In the two cases examined,
68 s c o t t w. d e s p o s a t o
informal institutions led to very different outcomes even in the context of identical
formal institutions. The implication is that explanations of political phenomena that
rest only on considerations of formal institutions will often be incomplete and lack
generalizability. Second, this chapter shows how the effects of informal institutions
can spill over beyond their local context. Previous research on informal institutions
in legislative politics has focused on the evolving rules and norms within the institu-
tion and their effects on legislation. This chapter, however, shows how informal
institutions outside the legislature can be just as important, shaping legislators’ basic
behavioral incentives. In particular, it shows that where the informal institution of
clientelism is present, politicians focus less on credit-claiming and more on delivery.
Where electoral markets are not constrained by the institutions of clientelism, politi-
cians’ priorities shift toward policy formation and credit-claiming.
In addition, this project suggests some revisions to purely institutional explana-
tions for political behavior. In particular, Brazil’s electoral system has been heavily
blamed for producing an ‘‘inchoate’’ party system. My analysis, however, suggests
that this particular set of institutional rules need not deterministically lead to feckless
democracy. The differences observed across states point to an important role for
noninstitutional variables.
More broadly, these results offer a challenge to any study that focuses exclusively
on formal institutions to explain political phenomena. The institutional arrange-
ments in Piauí and São Paulo are as close to a natural experiment as we could
observe anywhere, and were mostly exogenously imposed. The diverse outcomes
observed—some attributable to voting behavior and campaign styles, others to his-
torical paths—present an important challenge to institutional studies. The outcomes
show that institutional theories are clearly incomplete unless we incorporate elector-
ate and societal variables, and that these variables can have very powerful effects on
the political arena. Such societal variables may prove even more influential than
formal rules.
chapter three
destine legislative agreements with party leaders who passed economic reforms in
exchange for policy concessions, patronage, and ‘‘pork.’’
At the most general level, this chapter shows that informal institutions can play an
important role in explaining the gap between the outcomes predicted by standard
institutional theories and the observed patterns of policy change and economic
reform. More specifically, the chapter illustrates how ghost coalitions helped to
promote interbranch cooperation and to surmount the institutional rigidities im-
posed by formal (electoral and legislative) rules. Although presidential coalition
building has helped solve problems of multiparty presidentialism in several Latin
American countries, including Chile, Bolivia, and Brazil, Ecuadorian coalitions
differ in one important respect: they tend to be hidden from public view. Publicly
visible cross-party coalitions are unpopular among Ecuadorian voters, who have
become suspicious of the nature of political deals. Thus party leaders view public
coalitions as nonviable politically, and negotiations between opposition parties and
the government must be clandestine. Here, the absence of roll-call voting mecha-
nisms in the legislature is crucial, for it allows vote-trading while protecting the
electoral reputation of coalition partners who vote for unpopular economic reforms.
At the same time, threats of ‘‘going public’’ with such arrangements provide an
especially powerful enforcement mechanism to deter noncompliance.
The first section of this chapter describes in more detail the puzzling approval of
economic reforms despite rigid political institutions and public disbelief. I then
explore the informal nature of ghost coalitions, highlighting various incentives and
sanctions. The third section illustrates the crafting of ghost coalitions by focusing on
the adoption of modernization reforms during the administration of President Sixto
Durán Ballén (1992–96). I conclude with a discussion of implications for the study of
informal institutions and coalition making in comparative perspective.
Like many other Latin American countries in the 1980s, Ecuador confronted the
challenge of replacing the state-led import-substitution industrialization model with
tight fiscal discipline aimed at servicing lofty debt commitments and promoting
economic liberalization. For more than two decades, Ecuadorian presidents battled
a highly fragmented Congress and a skeptical public to enact adjustment policies
and structural reforms recommended by international financial institutions, in order
to become eligible for fresh international credits. Such policies, largely shaped by
the so-called Washington consensus, prescribed the adoption of financial and trade
liberalization, fiscal discipline, reduction of the public sector (including privatiza-
legislative ghost coalitions in ecuador 71
tions), and flexible labor reforms (Williamson 1990, 402; Hey and Klak 1999, 67).
During this period, Ecuador’s reform attempts were also affected by a series of
powerful exogenous shocks, including fluctuations in international oil prices, armed
conflicts, and natural disasters.≤ In the face of low growth, widespread unemploy-
ment, galloping inflation, a rapidly depreciating currency, and a severe financial
crisis, the economy teetered on the brink of collapse.
From the standpoint of existing institutional analyses, several aspects of Ecuador’s
formal institutional framework should have made successful policy reform unlikely.
First, the use of a closed-list proportional representation system to elect legislators
between 1979 and 1996 and the absence of effective barriers for obtaining congressio-
nal representation increased the effective number of parties and prevented presi-
dents from obtaining partisan majorities in the legislature. Second, the peculiar
combination of term limits with midterm elections seemed to remove incentives
among legislators to make long-term agreements.≥ Third, partly to confront such
congressional opposition, Ecuadorian presidents are endowed with some of the
strongest constitutional powers in the region, including strong legislative powers to
decree and veto legislation. In this context, there seemed to be few political incen-
tives to ensure the political cooperation and power-sharing necessary to pass reform
legislation.∂
The media and public opinion exacerbated the problem of making coalitions for
reform. As in other Latin American countries, government efforts to enforce fiscal
discipline and promote economic liberalization were often confronted by public
discontent and mobilization. Voters felt betrayed by parties and politicians who
attempted market reforms despite campaign promises, and their discontent was
reflected in surveys (Stokes 2001). Job approval ratings between July 1988 and March
2002 (figure 3.1) show that Ecuadorian presidents reached (net) negative ratings
before six months in office, and by the end of their first year in office, they had lost
well over 40 percentage points in popularity (Araujo 1998).∑
Widespread popular discontent with government performance thus should have
hindered the president’s ability to recruit coalition partners for passing economic
reforms. Indeed, political parties often publicly adopted an anti-gobiernista (anti-
government) discourse to dismiss any association with the government and to assert
their independence from presidential initiatives, especially in the vicinity of new
elections. In this context, collaborating with a generally unpopular president ac-
quired a pejorative connotation. According to Congressman Wilfrido Lucero, when
his Democratic Left (ID) party attempted to negotiate a tax reform with President
Jamil Mahuad in 1999: ‘‘We were disqualified, accused of being gobiernistas, by
other parties and our own. This has been a political prejudice (complejo politico),
72 andrés mejía acosta
100
80
(Favorable-Unfavorable) Job Approval Rates
60
40
20
-20
-40
-60
-80
-100
0 6 12 18 24 30 36 42 48 54
Number of Months in Office
Net Support Regression Fit
Figure 3.1. Presidential Net Job Approval Ratings (percent): Quito and Guayaquil (July 1988–
March 2001). Source: Informe Confidencial, Quito.
some kind of bad word in the political arena when it comes to negotiating with the
executive.’’∏ As former President Mahuad himself put it: ‘‘The worst insult you can
tell an Ecuadorian congressperson is to be a gobiernista.’’π
The media contributed to raising the costs of government cooperation by asso-
ciating incidents of interbranch bargaining with illegal or corrupt practices. When
asked about the role of the media in the formation of government alliances, eleven of
twelve legislators interviewed in 1999 claimed that the media made government
cooperation less likely, since media had a tendency to demonize political alliances
by focusing on corruption scandals rather than positive actions of Congress.∫ Over
the years, popular slang embedded this anti-gobiernista discourse in the Ecuadorian
political culture.Ω In this sense, public mistrust and the media’s role in disclosing
political alliances severely increased the politicians’ costs—and electoral liabilities—
associated with crafting public cooperation agreements with the government.
Notwithstanding this difficult environment, Ecuadorian governments managed
to avoid the hyperinflation crises and control the spiraling fiscal deficits that plagued
other Latin American economies in the 1980s and early 1990s. Presidents obtained
legislative ghost coalitions in ecuador 73
the approval of legislation to improve tax collection and limit government spending
during the 1980s (Grindle and Thoumi 1993). Under the generic label of ‘‘modern-
ization reforms,’’ the government passed significant financial, trade, and banking
liberalization laws in the early 1990s, despite having little or no formal party support
in Congress (Hey and Klak 1999; De la Torre et al. 2001). In 2000, a coalition of
opposition parties in Congress passed the government-proposed Monetary Stability
and Economic Recovery Law (Trole I) to adopt the U.S. dollar as the national
currency.∞≠ Other important reforms such as the Fiscal Responsibility, Stabilization
and Transparency Law (approved in March 2002), as well as financial, banking, and
labor legislation required to implement dollarization reforms, were passed despite a
minority presence of the government party in Congress.
From a comparative perspective, the institutional barriers to successful policy
reform seem to have proven far less daunting in Ecuador than one might predict.
Most notably, Ecuador’s General Index of Structural Reform through 1995 (.801) was
only slightly below the regional average (.821).∞∞ Indeed, Ecuador liberalized to a
greater degree than countries with far less party fragmentation, including Mexico
(before 1997) and Venezuela (before 1993).
The question, then, is why, given the country’s fragmented institutions and ad-
verse public opinion, were Ecuadorian governments consistently able to ‘‘muddle
through’’? The next section develops one set of answers: ghost coalitions, an informal
institution adopted by Ecuadorian presidents and party leaders to overcome the
constraints imposed by a multiparty presidential system and enable the approval of
economic reforms.
Students of Latin American politics are well aware of the instrumental role
played by clientelistic practices in cementing political agreements between different
agents, whether politicians seeking the support of potential voters, or party leaders
trading favors. At the same time, most accounts stress the inefficient and transient
nature of clientelistic practices for achieving determined outcomes, claiming that
they solely favor opportunistic and expensive instances of cooperation over finite
issues.∞≤
The Ecuadorian experience of ghost coalition formation suggests an alternative
view. Treating the informal system of patronage associated with such coalitions as a
specific form of clientelism (Mainwaring 1999, 176),∞≥ I argue that it provides actors
with the incentives and punishments to sustain cooperation games over time, thus
enabling them to achieve legislative outcomes that would not be possible through
74 andrés mejía acosta
formal institutions alone. Such clientelistic exchanges are based on mutually shared
expectations in which political elites believe that the failure to comply carries some
kind of external sanction, or at least the exclusion of certain agents from the benefits
of the political trade. Ghost coalitions may be viewed as complementary informal
institutions (Lauth 2000; Helmke and Levitsky, introduction to this volume), for they
enhanced governability by ‘‘filling in the gaps’’ and improving the efficiency of the
legislative process.
Despite the clandestine character of ghost coalitions, observers of Ecuadorian
politics have been aware of them since the return to democratic rule. As any newly
elected Ecuadorian president soon comes to realize, vote-buying (trading legislative
votes for political favors) is ‘‘the name of the game’’ for crafting policy coalitions in
Congress.∞∂ As in many other Latin American democracies, Ecuadorian presidents
have discretion over the distribution of a wide range of public ‘‘goodies,’’ including
appointments and particularistic concessions to potential political allies, such as
cabinet positions, subnational authorities, granting of licenses and contracts, alloca-
tion of government resources, and other payments.∞∑ What is distinctive about the
Ecuadorian case is that the favored parties often need to protect themselves from the
public censure of being called gobiernistas. Thus, the specific mutual understanding
underlying ghost coalitions between presidents and congress can be expressed as
follows: (1) presidents are expected to distribute clientelistic payoffs in exchange for
legislative cooperation, and (2) maintaining the secrecy of an agreement is a precon-
dition for effective legislative cooperation. The first premise refers to the accepted
currency for enabling coalition formation; the second refers to the nature of the
transaction itself.
Concealing political agreements from the public sphere allows politicians to
benefit from the gains of the trade while protecting parties from electoral liability for
collaborating with the government. During the Hurtado administration (1981–84),
for example, Gary Esparza was elected vice president of Congress in 1982 and
president in 1983 with the support of the Christian Democrats (Popular Democracy;
DP), the Democrats (PD), and his own Roldosista Group (GR). Although the al-
liance had been in place for nearly two years, ‘‘no party dared to call this relationship
a ‘pact’ much less a government pact’’ (Mills 1984, 87).∞∏ Indeed, Esparza—himself a
GR legislator—consistently denied being the candidate of the official alliance in
1983, and when reporters asked about his decisive collaboration in passing past key
government proposals, Esparza replied that his ‘‘only interest was the preservation of
democracy [from a military reversal] and not any government cooperation.’’∞π Dur-
ing the Mahuad administration (1998-2000), the media reported on the workings of a
‘‘legislative steamroller,’’ or aplanadora, to describe a perceived association between
legislative ghost coalitions in ecuador 75
the governing DP party and the Social Christian Party (PSC). Party leaders on both
sides deny the existence of a political alliance between the two, but this ‘‘coinci-
dence of interests’’—as they prefer to call it—facilitated the joint appointment of the
attorney general, the ombudsman, and the directory board of the Electoral Tribunal
(with the PSC in the presidency and DP in the vice presidency). In Congress, both
parties supported the approval of a peace treaty with Peru, the approval of a major
fiscal reform package (including the creation of new taxes) to finance the deficit, and
the 1999 budget. President Gustavo Noboa (2000–2003) helped cement a ghost
coalition with the appointment of ‘‘independent’’ politician Patricio Jamriska to the
Health Ministry in January 2001.∞∫ Although all sides denied any formal collabora-
tion, Jamriska’s ties to the National Integration Movement (MIN) help explain that
party’s crucial involvement in supporting subsequent government initiatives, includ-
ing key fiscal reform packages, in 2001 and 2002.
A crucial element for making ghost coalitions work in Ecuador is that there is no
public record of legislative votes. Roll calls are rarely taken in Congress and, over
time, legislators have consistently blocked the implementation of electronic vote-
counting mechanisms on the floor.∞Ω As one legislator put it, secret voting gave
legislators more flexibility for crafting alliances: ‘‘this way they claim to oppose
something but vote in favor of it.’’≤≠ I further explore the logic of ghost coalitions by
focusing, first, on the trade-offs posed by various types of patronage and, second, on
the mechanisms used to monitor and enforce compliance with such agreements.
perks in exchange for legislative support. Similarly, the Noboa government compen-
sated for its lack of partisan support by offering coalition rewards to the ‘‘indepen-
dent’’ MIN, mostly composed of disgruntled DP members. This pattern of making
informal coalitions with proto-parties composed of independent legislators chal-
lenges the conventional view that Ecuador’s so-called independent legislators are
‘‘loose’’ or ‘‘floating’’ (Conaghan 1995).
Because ghost coalitions are clandestine rather than public, monitoring and
sanctioning are critical issues for ensuring effective cooperation between presidents
and party leaders. Acting in an institutional environment that offers few formal
incentives for alliance formation, and often cementing alliances through unsavory
and even illicit deals, coalition partners cannot turn to officially sanctioned channels
to enforce compliance. However, political actors employ a wide range of levers and
strategies, ranging from judicial investigations and impeachment threats to party-
sponsored street protests and other regional and municipal conflicts,≥∂ to try to
enforce compliance with legislative agreements.
For presidents, the most important sanctioning mechanism lies in discretionary
power to dismiss appointed officials affiliated with parties that break the terms of the
alliance. The use of this authority to reward or punish congressional coalition part-
ners is clearly seen in the case of governors, who in Ecuador are appointed rather
than elected. Available data on the duration of mandates of provincial governors
indicate an extremely high turnover rate: governors lasted an average of 1.2 years, and
only 28 percent of those who left the office did so due to ‘‘natural’’ causes (comple-
tion of mandate or death). Leaving aside those governors who were ‘‘promoted’’ or
transferred to another posting, it can be inferred that more than 53 percent of
governors were directly dismissed by the president or were asked to resign.≥∑ As in
cabinet dismissals, presidents removed (or threatened to remove) appointed gover-
nors to ensure compliance with agreements made with congressional partners or to
accommodate new coalition partners. Reshuffling of provincial authorities is also
associated with periods of successful policy change.
Political parties use a variety of sanctioning mechanisms to ensure presiden-
tial compliance with the terms of ghost coalitions. For one, legislators may pun-
ish presidents by requesting impeachment proceedings against cabinet members
and censuring them with a simple majority of votes.≥∏ It is often argued that
the congressional opposition in Ecuador consistently dismantled cabinets through
legislative ghost coalitions in ecuador 79
70
64
60
50 49
50
Number of Occurrences
40
30 27
20 15
10
10 6 7
5
2 3 3
0
1979-1984 1984-1988 1988-1992 1992-1996
Administrations
Threat of Impeachment Impeachments Dismissals
Figure 3.2. Threats versus Actual Impeachment Proceedings across Four Administrations
(1979–96). Source: Sánchez-Parga (1998, 85, 203).
(Hey and Klak 1999). Durán Ballén inherited a moderately growing economy (3.6%
annual GDP increase) with a relatively small fiscal deficit of 1.2 percent of gross
domestic product.
Durán Ballén’s administration was marked by a dramatic lack of partisan support
in congress. Before the 1992 presidential election, Durán Ballén had abandoned the
PSC to form his own party, the Republican Union Party (PUR). Although the move
helped him win the presidency in 2002, the PUR gained only twelve of seventy-seven
seats in the legislature, which, combined with the six seats of his Conservative Party
(PCE) allies, gave the government control over less than a quarter of the congress.
Confronted with this scenario, the new president opted to seek a legislative agree-
ment with the PSC, the largest party in Congress (with twenty-one seats). Notwith-
standing Durán Ballén’s defection in 1991, the PSC was a natural coalition partner,
since it shared a center-right ideology and a similar commitment to market-oriented
reforms. An alliance with the governing party created a political dilemma for the
PSC: on the one hand, party leader Jaime Nebot had been a runner up in that year’s
presidential election, and he did not want to be seen collaborating with the govern-
ment he was planning to run against in 1996. On the other hand, the PSC sought
access to state resources, and party leaders were interested in passing economic
reforms that would facilitate their own policy goals in the future.
Government officials, including Vice President Dahik and Chief of Staff Mar-
celo Santos, held a clandestine meeting with PSC leaders León Febres Cordero and
Jaime Nebot, at a private Guayaquil residence, to craft a legislative agreement. The
context and contents of the agreement were only revealed three years later, when
Vice President Dahik made it public during his congressional impeachment hear-
ings (Congreso Nacional del Ecuador 1995). The ‘‘broad political agreement’’ with
the PSC required the government to make budgetary allocations for PSC-controlled
electoral strongholds in the coastal provinces of Guayas, Manabí, Los Rios, and
Esmeraldas, totaling more than 200,000 million sucres (approximately US$100 mil-
lion). It also included direct cash transfers—from the government’s discretionary
spending fund (gastos reservados)—to purchase individual legislators for as much as
US$500,000 each to approve laws on the privatization of telecommunications or
electricity, government jobs for PSC cronies in the state-owned oil (Petroecuador)
and energy (INECEL) companies, and, most importantly, the PSC’s control of the
Supreme Court and the Electoral Tribunal.∂≥
The government also used particularistic payments (usually from discretionary
funds) to purchase the support of smaller parties and independent legislators to
secure the required legislative majorities.∂∂ These negotiations included the ill-fated
written pact (mentioned above) with the National Concentration Legislative Bloc,
82 andrés mejía acosta
exile to Costa Rica when a Supreme Court justice issued a warrant for his arrest
(Sánchez-Parga 1998, 121; Saltos Galarza 1999, 209).
In the aftermath of the vote-buying scandal, several constitutional reforms were
passed to curb the discretionary spending funds of the executive and to limit the
abilities of legislators to negotiate resource allocation for their provinces. The PSC
did not escape unharmed from the corruption scandal, and its presidential candi-
date, Jaime Nebot, lost the 1996 presidential election for a second time.
conclusion
Crafting ghost alliances with the opposition has been an informal but systematic
practice employed by political leaders to surmount the institutional rigidities of the
Ecuadorian system and obtain significant policy reforms in the legislature. Accord-
ing to an informal set of rules for coalition formation, policymakers shared the
common expectation that legislative votes could be traded for particularistic pay-
ments and policy concessions. But making policy agreements with usually unpopu-
lar presidents imposed high liability costs for potential partners, especially before
electoral events. Thus, the secrecy of such agreements was an important part of the
informal rules of coalition formation, protecting politicians from the electoral im-
plications of backing unpopular economic reforms by the government. Yet, the threat
to break the ‘‘secrecy clause’’ and eventually go public with a clandestine agreement
was also an instrument for monitoring compliance: disgruntled politicians issued
different kinds of warnings to promote compliance with existing agreements.
The construction of informal elite power-sharing agreements, of course, has been
common among Latin American democracies (Siavelis, this volume). What makes
the Ecuadorian case distinctive is the fact that political currencies for coalition
making have been curtailed or demonized by the media. In this context, party
leaders forged clandestine coalitions with the government while at the same time
publicly denying any association with it—and even actively proclaiming their op-
position in order to maintain an image of ‘‘political chastity.’’ The absence of roll-call
voting and legislative watchdogs allowed the survival of this informal institution.
This chapter offers two contributions to the study of political institutions in Latin
America. First, it goes beyond the description of the formal institutional framework
and its predictions for policy change. By analyzing the informal rules of coalition
making, I explore the value and nature of different coalition incentives and the
effective use of threats and sanctions to enforce legislative cooperation. In other
words, this study avoids focusing on the purely written rules in favor of exploring the
84 andrés mejía acosta
‘‘actual rules’’ guiding coalition building in Ecuador. In so doing, it opens the door
for future comparative research. One potential area of research, for example, lies in
comparing informal coalition dynamics in Latin American presidential democracies
and in parliamentary minority governments in Europe.
A second contribution has been to outline how the interaction of formal and in-
formal institutions can shape the prospects for democratic governance. I have argued
that ghost coalitions played a complementary role by helping to improve the (poor)
policymaking capacity of a fragmented legislature. By enabling executives to build
functioning legislative majorities through the distribution of pork, patronage, and
policy concessions, ghost coalitions facilitated governability, which helped Ecua-
dor’s fragile democracy survive some of the worst economic crises in its history. From
the perspective of accountability, however, such patterns of clandestine cooperation
also hindered the development of a policy-oriented or programmatic connection
between parties and voters at the national level. Thus, from a normative perspective,
informal institutions have ‘‘double-edged’’ effects, playing both positive and negative
roles simultaneously (Helmke and Levitsky, introduction to this volume).
This analysis thus reveals the inherent tension in institutional reforms designed to
‘‘clean up’’ Ecuadorian politics.∂∫ Recent efforts to reform the political system have
focused on curbing the available currencies that entice legislative cooperation, such
as doing away with the executive’s use of discretionary spending accounts, eliminat-
ing legislators’ ability to negotiate budgetary allocations for their provinces, or pun-
ishing with mandate recall those legislators who defect from their parties or bargain
for government jobs.∂Ω Though beneficial from the standpoint of transparency and
accountability, such reforms could well have the (largely unintended) effect of
undermining Ecuador’s fragile governability.∑≠ As one party leader put it, eliminat-
ing those bargaining chips is like ‘‘tying our hands and feet’’ when it comes to making
alliances.∑∞ Understanding the informal rules and dynamics of existing vote-buying
practices, in a context in which the ‘‘difficult combination’’ of multiparty presiden-
tialism prevails, may well be a starting point to improve the notoriously fragile bases
of Ecuador’s governability.
part ii
Informal Institutions
When Formal Contracting Is Prohibited
Campaign Finance in Brazil
david samuels
zilian experience in perspective and suggest how the theoretical framework might be
more broadly applied.
≤ H1: The overall supply of campaign funds is a function of the incentives that
potential financiers have to influence the distribution of government services.
≤ H2: The overall demand for campaign funds is a function of the degree of
inter- and intraparty competition.
≤ H3a: Market exchange will proceed to the degree that politicians desire to
establish collective and/or individual reputations.
≤ H3b: Market exchange will proceed to the degree that politicians and donors
expect to engage in repeated interactions.
≤ H3c: Market exchange will proceed to the degree that politicians and donors
have established quasi-institutional sanctioning mechanisms.
Under certain conditions, demand for campaign funds may be low. For example,
campaigns may be wholly or partly publicly funded, eliminating or reducing candi-
dates’ ability or need to raise funds. When campaigns are publicly funded, even
informal campaign finance contracts should not exist, because (at least in principle)
private sources of money are not relevant to the electoral process.∂
When campaigns are privately funded, under what conditions will demand exist?
The main factor is the degree of electoral competition. On the one hand, demand
may be low because campaigns are cheap, because they are not competitive. If one
candidate or party has scared off viable rivals, there would be little need to raise
money because there would be no need to spend money to win votes. On the other
hand, as the competitiveness of elections increases, so do their costs, and so does
politicians’ demand for money (Crain 1977). Such demand is also a function of the
particular electoral rules. For example, in countries with rules such as Brazil’s open-
list system, the demand for campaign funds is a function not only of interparty
competition but also of the degree of intraparty competition in each constituency
(Cox and Thies 1998). That is, if candidates must compete against one other for votes
in each district, then they have strong incentives to spend money in an attempt to
differentiate themselves from their copartisans as well as from candidates of other
parties.∑ Whether intraparty competition matters depends on the electoral system.
Thus, when candidates face competition, we need only assume that they desire to
win to conclude that a demand for campaign funds exists. The degree of demand
depends on intra- and interparty competition and other costs of running a campaign,
which will vary widely based on contextual factors.∏ However, other elements are
required to ensure credible commitments and thus the generation of informal cam-
paign finance institutions.
Supply and demand are necessary but insufficient reasons for market exchange to
proceed and thus insufficient for informal institutions to emerge. Market exchange
campaign finance in brazil 93
also requires contracting between those who produce the supply and those who
generate the demand. Contracting is problematic, because both buyers and sellers
have few guarantees that the terms of the deal will be fulfilled. For example, poten-
tial campaign financiers may be reluctant to invest because they lack confidence
that they will get what they paid for. This issue boils down to whether suppliers of
campaign funds expect politicians to provide an adequate return on money ‘‘in-
vested.’’ If they do not, then campaign funds should be undersupplied. Even if
investors are interested in obtaining government services and politicians demand
money, we may see underinvestment because of contracting problems—that is,
because informal institutions have failed to emerge. By contrast, if investors are
confident they will obtain a good return on their investment, then we have good
reason to suppose that campaign funds will be supplied.
What guarantees do potential contributors have that politicians will come through
with their end of the bargain? Unfortunately, campaign financiers can never obtain a
rock-solid guarantee. The reason is quite straightforward: in any country, campaign
funds contracts are legally unenforceable (McCarty and Rothenberg 1996; Kroszner
and Stratmann 1998). Contributors cannot take their case to a higher authority if
a politician ‘‘reneges on a deal,’’ nor can a politician sue contributors for failing
to provide ‘‘promised’’ campaign funds. Consequently, credible commitments in
campaign finance contracts anywhere cannot rest on institutionalized punishment
mechanisms.
Given the absence of credible sanctioning mechanisms, contributors’ invest-
ments may be subject to ex post expropriation without compensation. In this situa-
tion, we might see underinvestment even where supply and demand are high. That
is, even if we assume that politicians do not actually ‘‘expropriate’’ contributions ex
post, the absence of formal contractual obligations to provide service in exchange for
money means investing in campaigns is at a minimum a highly risky proposition.
What, then, maintains the credibility of campaign funds contracts? The literature
on credible commitments provides clear guidelines as to the roots of secure contract-
ing: when transaction costs are high, political actors (both ‘‘buyers’’ and ‘‘sellers’’ of
campaign funds) have incentives to develop institutions or quasi-institutional mech-
anisms to counteract the effects of uncertainty and enhance the possibility of suc-
cessful contracting. Political actors can develop three kinds of mechanisms to en-
hance the credibility of their commitments: reputation, repetition, and sanctions.
Consider reputation first. To enhance their credibility, politicians have incentives
to develop a ‘‘brand name.’’ A successfully developed brand name indicates quality to
potential voters, and also provides voters with a target against which to retaliate if
quality does not meet expectations (Akerlof 1970, 500). Candidates could attempt to
94 david samuels
tributors expect to repeatedly interact with politicians over a relatively long period,
iteration can help establish credible commitments.
Still, even if we assume that politicians want to build their reputations and have
long-term career goals, North and Weingast (1989) have argued that reputation and
iteration are insufficient to support institutionalized political markets. Thus we can-
not yet conclude that we have ‘‘enough’’ to generate informal campaign finance
institutions. This is because neither reputation nor iteration can fully restrict politi-
cians’ ex post behavior or prohibit them from manipulating the market in their favor
once they are in office. This occurs only when threats to punish defection are made
credible, through the creation of sanctioning mechanisms. However, as noted above,
such institutions are precluded in all campaign funds markets, at least formally,
because the state will not enforce implicit campaign finance deals.
We might therefore conclude that because campaign finance contracts are un-
enforceable, a campaign finance market does not exist, anywhere, and thus informal
institutions in campaign finance markets are chimerical. Or we could hypothesize
that sanctioning mechanisms can also emerge informally. Because both contributors
and politicians have strong interests in maintaining exchange in the campaign fi-
nance market, they could endogenously develop enforcement mechanisms or be-
havioral norms to enhance credible commitments. Scholars have demonstrated that
such mechanisms have emerged in the United States: politicians have deliberately
crafted quasi-institutional contractual safeguards such as seniority rules and commit-
tee specialization as ways to enhance their reputation and the likelihood of repeat
play (Kroszner and Stratmann 1998, 1183).
Campaign funds contracts cannot rest on formal institutional safeguards. Instead,
exchange must rely on informal institutions of reputation, iteration, and sanctions.
To illustrate these hypotheses more clearly, I now turn to the case of Brazil.
Until 1994 we lacked any hard information about the cost of Brazilian campaigns,
because only in that year did candidates have to start submitting records of all
contributions (e.g., Samuels 2001a, 2001b, 2001c, 2002). We now possess data con-
firming that elections at all levels in Brazil are very costly, and that money matters a
great deal.π Tables 4.1 and 4.2 provide information on the amounts of money candi-
dates raised for campaigns in Brazil.∫
Brazilian elections are expensive, across the board. To put these costs in perspec-
tive, we can first note that in August 2002 the U.S. dollar was worth about R$2.50.
table 4.1
Declared Contributions, Presidential Races, Brazil
Contributions
Year Candidate (R$, Aug. 2002) Vote %
table 4.2
Average Declared Contributions by Office
Contribution (R$, Aug. 2002)
Year Position (no.) Winners Losers All
1994 Governor (82) 5,929,190 1,523,653 2,680,107
(87.5%) (62.8%) (67.8%)
Senator (127) 857,311 316,739 485,268
(91.2%) (53.8%) (62.7%)
Deputy (1,542) 252,401 65,139 114,808
(90.9%) (53.4%) (59.9%)
1998 Governor (81) 3,263,717 1,094,682 1,549,912
(62.9%)
Senator (90) 1,392,398 106,475 477,964
Deputy (1,849) — — 101,884
(54.1%)
2002 Governor (145) 2,762,693 899,862 1,223,270
(96.3%) (71.3%) (74.7%)
Senator (242) 814,926 149,223 295,018
(98.1%) (72.5%) (77.3%)
Deputy (2,554) 224,433 36,254 74,034
(100%) (55.0%) (60.5%)
Note: Percentages in parentheses are percentage of candidates in each cell who provided data.
campaign finance in brazil 97
Thus, for example, even though a winning congressional candidate in Brazil does
not spend nearly as much as a winning U.S. House candidate does ($530,000 on
average in 1994), candidates in Brazil are legally prohibited from purchasing tele-
vision or radio advertising,Ω and a dollar goes a lot further in Brazil, where annual per
capita income is less than US$5,000.
Clearly, a market for campaign funds thrives in Brazil. Why is this so, and what
supports informal campaign finance institutions? Why do potential contributors
provide money? Potential contributors in Brazil are very willing to open their wallets
because elected officials control the distribution of valuable resources—for example,
export subsidies, banking regulations, exchange rate controls, and pork-barrel funds
(Ames 1987, 1995a, 1995b; Abrúcio 1998; Bezerra 1999). Let me focus briefly on the
exchange of campaign funds for government public works contracts. Brazilian legis-
lators can influence the geographic distribution of pork-barrel projects, and spend a
good deal of their time attempting to do just that (Ames 1995b). Construction com-
panies desire government contracts for such projects, and they have two reasons to
provide a supply of campaign funds to politicians. First, they want the funds directed
to where their firm operates, as opposed to where other firms do. They invest in
candidates who, so they believe, will successfully ‘‘bring home the bacon’’ (Bezerra
1999, 242). Second, politicians not only influence the distribution of projects and
authorize their initiation, but can also influence when and even whether the firm
will actually get paid for an existing project (Bezerra 1999, 245). Because funds for
pork-barrel projects are scarce, potential donors have strong incentives to invest
heavily, so that they are not excluded from the division of the spoils.
An exploration of the demand side of the equation also explains why Brazilian
campaigns are so costly. Demand is a function of inter- and intraparty competition,
and both forms of competition in Brazil have grown increasingly fierce since the
return to competitive elections. The number of effective parties in legislative elec-
tions rose from 2.4 in 1982 to 8.2 by 1994 and has stayed fairly high since then
(Nicolau 1997, 316). Candidates in all parties now face more intense interparty
competition. Because party organizations are generally weak and do not help indi-
vidual candidates on the campaign trail, candidates must raise their own funds to
survive the electoral battle—and must raise as much as they can.
98 david samuels
We know that campaigns are expensive in Brazil partly because of the consider-
able incentives to provide supply and generate demand. However, these are insuffi-
cient for market exchange to occur and thus insufficient to explain the emergence of
informal campaign finance institutions. We still must discover whether reputation,
iteration, and sanctions exist to support such institutions. Scholars concur that Bra-
zilian politicians—even those on the left, though to a lesser degree—strive to con-
struct a strong personal reputation for delivering particularistic goods (Ames 1995a,
1995b; Mainwaring 1998). Politicians seek recognition for generating public works
project for their region, and for providing individualized favors such as jobs, food, or
small loans. Such a reputation is the key to most Brazilian politicians’ career success,
and few politicians survive in the system merely as advocates of programmatic policy
positions. A Brazilian politician who turned away from the provision of particularis-
campaign finance in brazil 99
tic goods would be subject to retaliation. A reputation as one who ‘‘brings home the
bacon’’ not only helps solidify a reputation with voters, but also helps the politician
develop a reputation with potential campaign contributors, who might also shy away
from a purely ‘‘programmatic’’ politician.
One might question whether firms and congressional politicians in Brazil can
develop long-term relationships, because turnover in the Chamber of Deputies
with each successive election exceeds 50 percent. In contrast, turnover in the U.S.
House is less than 10 percent with each election. Do Brazilian deputies not ‘‘stay
bought’’? In fact, Brazilian politicians do have long-term political careers, even
though they may not have long-term congressional careers. Most deputies arrive
in the Chamber after spending some years serving in a state legislature, on mu-
nicipal councils, or even as mayor. Typically they spend one or two terms in the
Chamber and then continue their careers outside the national legislature, return-
ing to state or local politics. For example, they seek to serve again as mayor or to
move to a more powerful position in state government (Samuels 2003). In Brazil,
executive positions in subnational governments wield significantly more power to
hire and fire and to implement public works projects than do positions in the
national legislature.
As in the United States or Japan, the structure of political careers in Brazil
solidifies politicians’ reputations and establishes that they will engage in repeat play
with potential campaign financiers, but in a particular way. Brazilian politicians
never stray far from their local or state origins, even while serving in the national
legislature; they begin cultivating relationships with campaign financiers early in
their careers. When they reach the Chamber of Deputies, they continue to develop
these relationships and seek to deliver more substantial government contracts. Sub-
sequently, their career path both leads them to positions with greater capacity to dip
into the pork barrel and returns them to where the projects will actually be imple-
mented. Firms with budget interests take a long-term view and seek to invest in
enduring relationships with politicians who will exert long-term influence over gov-
ernment resources (Bezerra 1999, 245).
Brazilian politicians have long-term political careers that require constant sources
of funding, and thus they have strong incentives to develop long-term relationships
with potential campaign financiers. Likewise, potential financiers want to support
politicians who will enrich their coffers through delivery of government contracts
over the long term. The structure of political careers in Brazil thus serves as an
analogue to the seniority system in the U.S. House and lends credence to the notion
that reputation and iteration support informal contracting mechanisms in Brazil
(Kroszner and Stratmann 1998, 1183).
100 david samuels
members of the candidate’s family or are the candidate’s business partners, as well as
the high probability that the politician plans to return to local politics later on—
serves to strengthen ‘‘sanctioning’’ norms. Specific examples from two states illus-
trate the situation.
In the state of Rio Grande do Norte, known for its patrimonial and insular politics
(not unlike many of Brazil’s other states), nine of the twenty-nine donors to one
winning deputy candidate in 1994 had his family’s name. These donors were respon-
sible for providing 85 percent of the candidate’s funds. Three of the seven individual
donors to another winning candidate from the same state were family members, and
these donors accounted for 53 percent of his campaign funds (Brazil, Tribunal
Superior Eleitoral 2002).∞≠ This familiarity is not only a characteristic of Brazil’s
smaller, more backward states. São Paulo, Brazil’s largest and most economically
advanced state, has a population of about thirty million (about the size of Argentina),
and accounts for about 40 percent of Brazil’s gross domestic product, which is the
tenth-largest in the world (perhaps). Still, the state’s industrial base is highly concen-
trated, and despite its large population and relatively complex economic base, rela-
tively few people and companies enter the campaign finance market. Several of the
candidates best known for their business connections received contributions from
relatively few firms. Deputy Antônio Delfim Neto, who previously held the positions
of Minister of Finance, of Agriculture, and of Planning, received contributions from
a grand total of seventeen firms and three people. The firms provided 85 percent of
his total of approximately R$700,000 in contributions in 2002 (Brazil, Tribunal
Superior Eleitoral 2002).
Politicians in Brazil seek to establish iterated, long-term relationships with poten-
tial campaign financiers to develop their reputations. They want to establish them-
selves as competent providers of pork, both to voters and to the firms that thrive on
government contracts. Given that most campaign contributions come from family
members and a small number of firms, Brazilian politicians are much more likely to
have a tight personal relationship with their financiers. This familiarity helps solidify
a reputation, enhances the long-term prospects for the relationship, and reduces the
likelihood that the politician (or the contributors) will ‘‘defect.’’
Sanctioning Mechanisms
Campaign finance relationships in Brazil seem quite ‘‘cozy.’’ Not only does famil-
iarity breed confidence, but the limited number of contributors to each campaign
serves as an informal oversight and sanctioning mechanism. It is likely that those
who give to a campaign know who else has given (the data are available on the
102 david samuels
internet soon after the election), and given their small numbers and the possibility
that kinship connections exist between them, donors can observe and share informa-
tion about the politician’s behavior. We therefore ought not to see many politicians
reneging on their deals, because information about their behavior is relatively good
among their financial supporters. Still, despite the apparent coziness of campaign
funds relationships, the possibility exists that politicians and/or financiers could
renege on implicit campaign funds contracts. A family feud could erupt, or competi-
tion for government contracts could emerge between firms in a politician’s baili-
wick, which would allow a politician to break a contract with one firm if another
offered a greater payoff.
However, the threat of punishment for reneging does exist. For example, a firm
that fails to receive preferential treatment in exchange for a contribution can refuse
to supply additional funds for the politician’s next election. Elsewhere, I have found
a strong statistical relationship between deputies’ ability to obtain pork from the
executive branch and their ability to obtain campaign finance (Samuels 2002). This
provides (indirect, at least) support for the notion that informal contracts exist.
Politicians who fail to ‘‘bring home the bacon’’ will subsequently fail to reap substan-
tial campaign finance contributions.
Officeholders can also play hardball. Politicians control substantial government
resources that potential campaign financiers desire. If a politician directs a govern-
ment contract to a firm with an implicit understanding that campaign funds will be
forthcoming in exchange, but then receives none, the politician can retaliate in at
least two ways. As noted above, he or she could refuse to direct any other projects that
firm’s way or could attempt to delay payment on the firm’s existing contract. A delay
of several months or a year could substantially affect a firm’s cash flow.
Sanctioning should not occur very often; we should observe it only in extraordi-
nary circumstances. Detecting such instances of defection will be extremely difficult
for an academic observer, but a good example does exist: the impeachment of
President Fernando Collor de Mello in 1992. In 1989, Collor won Brazil’s first
democratic presidential election since 1960, but by early 1992 his administration was
embroiled in a scandal that had at its center his campaign funds ‘‘bag man,’’ Paulo
César Farias. In 1989, Farias had requested and obtained massive infusions of cash to
win the election against the leftist Workers’ Party candidate Lula, who led early in
the polls. However, after the election, Farias made the rounds once again and
demanded more cash, this time in exchange for ‘‘access’’ to the newly elected presi-
dent. ‘‘Going to the well’’ a second time violates the spirit of the campaign finance
contract, since it seems that donors gained nothing from their initial contribution
before Farias reappeared and demanded additional funds. Because the requests for
campaign finance in brazil 103
money now openly stank of extortion, many firms refused to provide additional
contributions. As punishment, Collor and Farias then ‘‘had [these firms’] disburse-
ments for goods and services already performed through 1989 halted, and they were
blacklisted from any new contracts.’’ Some of the more vulnerable firms were bank-
rupted (Fleischer 1997, 302).
This attitude violated accepted practice. ‘‘Normal’’ campaign finance contracts
carry implicit understandings that money buys at least access, if not influence. In
contrast, Collor’s scheme implied that money buys nothing, and failure to con-
tribute would result in government interference in your business. Moreover, Collor
apparently did not let sleeping dogs lie—he actively attempted to ‘‘undermine es-
tablished political forces and power centers’’ that opposed his scheme (Weyland
1993, 13). These forces were not just opposition parties, which lacked broad represen-
tation in Congress, but members of the traditional political class and the business
community.
Collor’s corrupt practices paralleled his attitude toward Brazil’s party system. He
sought to govern without Brazil’s established political parties and regional leaders
and their congressional representatives, refusing to engage in the pork-barrel ‘‘wheel-
ing and dealing’’ that had long characterized Brazilian politics. Thus Collor not only
failed to share the wealth through his corruption schemes, he alienated potential
political allies through his unilateral governing style. This also deviated from ex-
pected norms.
Collor’s presidency failed because the enormous scale of his corruption schemes
attracted substantial press attention, and because his corruption and his governing
style alienated the dominant, established political class. It remains unclear whether
these politicians wanted a piece of the action or wanted an end to the corruption.
Regardless, the point remains that Collor had violated expected norms of behavior.
(My view does imply that if Collor had ‘‘shared the wealth’’ a bit earlier, he could
have saved his presidency.) As his administration came under increasing press scru-
tiny, Congress opened an investigatory commission. By that time the business com-
munity and those politicians whom Collor had excluded or punished returned the
favor by refusing to support the president when he attempted to negotiate for their
support. In the end, Collor’s scheme came crashing down and he was impeached by
a Congress unwilling to let him rewrite the rules of political bargaining in Brazil.
Collor’s impeachment demonstrated not only that Brazil’s political institutions can
work as expected, but also that politicians will punish those who clearly divert from
long-established expected behavioral norms.
Even a relatively chaotic and fragmented political market like Brazil’s can sup-
port the ingredients necessary for credible campaign finance contracts to emerge.
104 david samuels
Politicians may have strong incentives to construct solid reputations that will endure
over time and that can develop sanctioning mechanisms. Politicians’ desire to de-
velop a ‘‘personal vote’’ characterizes many systems. Although Brazil’s particular
career path may be unique, politicians elsewhere are likely to have established
signaling devices or ways to provide contributors with the information that they are
in politics for the long haul. Likewise, the importance of family is not unique to
Brazil, nor is the highly elitist nature of politics. In short, in many countries, similar
dynamics may ensure that market exchange in campaign finance can occur and thus
informal campaign finance institutions may emerge that are accommodative to the
formal prohibition of bribery and extortion, yet reflect donors’ desire to influence
government and politicians’ desire for money to continue their careers.
conclusion
In this chapter I have outlined five hypotheses that permit a cross-national explo-
ration of the emergence of informal campaign finance institutions. First, the supply
of money in elections is a function of the incentives that potential contributors
have to influence the distribution of government services. Second, the demand for
campaign funds is a function of the degree of inter-and intraparty competition.
Third, market exchange and thus informal contracting will occur to the degree that
(a) politicians seek to establish a reputation, (b) politicians seek to engage in re-
peated interactions with donors, and (c) politicians and donors have established
quasi-institutional sanctioning mechanisms. Additional research is obviously neces-
sary to further assess these hypotheses and to further explore the importance of
campaign finance for politics.
These hypotheses may not hold everywhere. Potential contributors’ incentives to
provide a supply of money will vary based on elected officials’ (as opposed to un-
elected bureaucrats’) ability to influence the distribution of government services
and funds. The focus of contributors’ largesse may also vary, based on the executive-
legislative balance of power. In addition, politicians’ incentives to develop a repu-
tation, play an iterated game, and develop secure contracting mechanisms will also
vary, depending on the nature of political career structures both internal and exter-
nal to political parties. As implied in this analysis of Brazil, we may have to look
to the state or even local level to detect sanctioning and reputation-enhancing
mechanisms.
In broad perspective, the ‘‘democratic problem’’ in Brazil and similar countries is
not the absence of credible commitments, or their relatively weak institutionaliza-
tion, but their excess. Informal institutions are the glue that holds clientelistic rela-
campaign finance in brazil 105
michelle m. taylor-robinson
whether deputies attend to local community and constituent needs, and indicate to
deputies that performing such service is an important part of their job and their
contribution to the party’s next campaign effort. Because Costa Rican deputies
cannot be immediately reelected, the only way they can continue their political
careers is if their party wins the presidency, in which case they can be appointed to a
post in the executive branch. Thus, if they do not help their party to maintain strong
local support, they may be unable to continue their political career.
In many Latin American countries, it is assumed that because deputies serve their
party and its leaders, local interests are not represented in the national policy debate
(Shugart 2001). Informal institutions may help fill this local representation gap.
Parties may be able to use the informal institution of patron-client relations, par-
ticularly clientelistic networks attached to parties, to ‘‘represent’’ (some) local inter-
ests. Informal institutions—such as clientelism—that have long helped people get
things done may enhance, or substitute for, the performance of formal institutions
installed as part of a new democratic regime (Whitehead 2002; Helmke and Levitsky,
introduction to this volume). For example, parties that developed in a patron-client
environment may expect deputies to maintain the support of their clients so that the
party will retain its mass support in the next election.≤ If so, clientelism is a ‘‘substitu-
tive’’ informal institution (Helmke and Levitsky, introduction). Clientelism is ob-
viously not a perfect substitute, as clients of opposition parties are excluded from
representation, and programmatic representation is minimal. However, particularly
for poor people, for whom state patronage and local public goods are likely to be of
greater value than representation on national policy issues (Desposato 2001), clien-
telism creates a chance that their legislative patron will represent their interests when
their party is in power. This substitutive role of clientelism may help explain why an
apparently anachronistic informal institution persists even as procedural democracy
consolidates. To explore these questions, this chapter examines the legislative record
of members of the Honduran Congress.
The electoral systems and clientelism literatures rarely speak to each other. The
former is firmly grounded in the rational choice school, and thus intent on showing
how electoral institutions create incentives for certain types of behaviors by rational
actors ranging from voters to parties to politicians. The latter literature is based in
sociology and anthropology, and it focuses on explaining social relationships in
traditional societies; political scientists studying machine politics have adopted it
108 michelle m. taylor-robinson
(Banfield and Wilson 1965; Scott 1969). The two literatures have in common the
assumption that actors pursue the strategy that they think will best enable them to
achieve their goals.
The electoral institution literature assumes that rational reelection-seeking politi-
cians will work to please the party elites and constituents who control their re-
nomination and reelection chances (Fenno 1978; Cain et al. 1987). The rational
patron will provide enough services to clients to keep them from looking elsewhere
for help, but will not squander scarce resources that could be used to acquire more
clients (Graham 1990).
Electoral institutions create incentives for politicians to seek a personal or a
partisan vote.≥ As the party becomes increasingly important for winning elections,
politicians have an incentive to build the party’s reputation—to make the party label
a meaningful brand name for voters. Where a responsible party model applies, the
voter can choose the party whose platform is closest to her ideal point. Where parties
do not present competing packages of policies, but compete for voter loyalty at an
emotional level, the party label still frees the voter from having to know each candi-
date as an individual. In either case, the voter votes for a party, not for the candidate
as an individual. By contrast, where parties have less influence on how people vote,
candidates have an incentive to seek a personal vote by creating an individual image,
through advertising, credit-claiming, and position-taking (Mayhew 1974). Closed-list
proportional representation (PR) electoral rules, particularly where national party
leaders control ballot access and list position, produce a strong incentive to build a
partisan reputation. Politicians have an incentive to build a personal reputation in
open-list PR and preference voting systems, and where competitive primaries deter-
mine nominations.
Electoral strategy also varies with district magnitude (DM). In closed-list PR
elections, the incentive to seek a partisan vote should rise as DM increases. The
larger the number of candidates on a party’s list, the harder it is to distinguish oneself
from copartisans, since the voter has to vote for a party’s entire list even if he would
like to vote for a particular politician. However, when DM is small, politicians can
seek a personal vote along with a partisan vote, because an individual candidate can
stand out to voters (Shugart et al. 2005). In open-list PR elections, the incentive to
seek a personal vote should increase as DM rises, as it becomes more important for a
candidate to distinguish herself from copartisans and from candidates of other parties
(Ames 2001).
Carey cautions that ‘‘when legislators cannot be reelected, they have no direct
electoral incentives to build personal support among voters through particularism,
from caudillismo to democracy: clientelism in honduras 109
regardless of how the electoral system is organized’’ (1996, 103–4). This logic should
also apply when reelection is permitted but is unlikely.
In the U.S. Congress, members who experience a close election are expected to
perform more constituency service than those who win by a large margin (Fenno
1978; Johannes 1984). In PR systems, a similar incentive should exist for deputies
elected from a marginal position on their party’s list. They should perform more
constituency service as part of their effort to move up the list to a safe position to
protect themselves from a possible vote swing against their party.
In sum, the electoral institution literature points to several hypotheses about
when politicians would be expected to engage in constituency service work:
≤ H1: Politicians will be more likely to perform constituency service when they
are elected by open-list PR than when elected by closed-list PR electoral rules.
≤ H2: Politicians elected by closed-list PR electoral rules will be more likely to
perform constituency service as DM approaches 1.
≤ H3: Politicians will be more likely to perform constituency service as the
chance of reelection increases.
≤ H4: Politicians elected from a marginal position on their party’s list will be
more likely to perform constituency service than those elected from a safe list
position.
The clientelism and machine politics literature also has implications for the
behavior of elected officials. Clientelism can be defined as a mode of social stratifica-
tion in which clients and patrons are tied to one another in a mutually beneficial but
unequal exchange relationship (Lemarchand 1981). A political machine is a party
organization in which local political bosses are patrons who offer inducements to
clients in exchange for political support and votes (Banfield and Wilson 1965). Two
specific circumstances are conducive to political machines. (1) A candidate needs to
win votes to obtain office, but can win votes by delivering favors and tangible rewards
to people, rather than promises about policy. (2) Traditional deference patterns have
broken down, so that parties maintain the support of their voters by providing ‘‘con-
crete, short-run, material inducements’’ (Scott 1969, 1146). As new loyalties emerge
(e.g., class, occupational ties) and the state provides universal social welfare pro-
grams to attend to basic needs, people become more concerned about policy or
ideology, and the machine is no longer feasible (Banfield and Wilson 1965).
When these propitious circumstances exist, the machine specializes in the ‘‘polit-
ical distribution of public works through pork-barrel legislation and in the dispensa-
tion of jobs and favors through more informal channels’’ (Scott 1969, 1147–48). This
110 michelle m. taylor-robinson
why honduras?
The Honduran Congress offers a likely case for finding that long-standing infor-
mal institutions affect the performance of newer formal democratic institutions. Two
traditional political parties with clientelistic connections to voters dominate Hon-
duran politics. The country’s limited economic resources and poverty mean that the
state faces demands for services that exceed its capacity to respond. Formulaic,
needs-based criteria for state aid to communities and individuals would bankrupt the
state and break tight budgets imposed by international lending agencies. Clientelis-
tic connections provide a mechanism for allocating state resources, and allow the
party in government to use time-honored methods to reward supporters.∂ For exam-
ple, a person may need a deputy’s help to get a doctor’s appointment at a state
hospital, or a scholarship for schoolbooks. When a new government (party) takes
office, throngs of party supporters (clients) line up for government jobs. On the
opposite side, a community organization affiliated with the out-of-power party will
be passed over for state resources.
The Liberal (PLH) and National (PNH) parties have dominated Honduran poli-
tics since the beginning of the twentieth century.∑ They were established as tools
in the competition between the international banana companies operating in the
from caudillismo to democracy: clientelism in honduras 111
country, but became autonomous entities after the banana companies merged in
1929. The parties continue to compete for power and the access to state resources
that power provides. During their periods in office, both parties have been accused
of corruption and political favoritism, as state resources (such as jobs, contracts,
infrastructure projects) have been distributed to supporters, and opposition support-
ers have been excluded from the spoils.
Clientelism (among other factors such as lack of raw materials and isolation) is
often blamed for the lack of economic development in Honduras (Posas and del Cid
1983; Euraque 2000; Mahoney 2001). During the many brief civil wars in the nine-
teenth century, and the period of competition between the banana companies in the
early twentieth century, it was not rational for villages to invest in economic develop-
ment, as battles often destroyed their efforts. It was easier to support the winning side
in a conflict and obtain a share of the state spoils (Munro 1967, 126–32). A patron was
a necessity in this poor country made up of isolated communities, as the government
had no capacity to take care of the vast range of people’s needs. Economic and
political elites took on the role of providers and, in return, their clients supported
them in elections and civil wars.
This expectation that political caudillos would share state resources with their
supporters has continued in the current democratic regime. Honduran presidents
have difficulty pursuing policy initiatives, despite their strong partisan powers (Main-
waring and Shugart 1997), because they are busy attending to the promises of ‘‘pork’’
and patronage they made to get elected (Taylor 1996). Deputies also feel pressure to
attend to the needs of communities and individual supporters. As one deputy re-
marked when explaining why it is difficult to focus on legislating: ‘‘Honduran voters
do not understand that a deputy’s job is to legislate, and they do not value legislation
because you cannot eat a law.’’∏ Local party activists work to promote the electoral
fortunes of their faction leader. In return, they expect jobs, personal favors, and pork-
barrel projects for their communities. As elsewhere, however, clientelism in Hon-
duras is an unequal relationship, for it is very costly for a client to defect from her
patron. The client can switch allegiance to another party, but will receive benefits
from doing so only if the party wins the presidency, and if there are not many more
people with stronger connections to the patron/party who precede her in the line
for benefits.
The current democratic regime in Honduras was installed in 1982 after a decade
of military rule, but, in many respects, democracy brought little real change. The
traditional parties continue to dominate elections, control all branches of govern-
ment, and dole out state resources to their supporters. The Christian Democratic
Party (PDCH) and the Innovation and National Unity Party (PINU) won a few seats
112 michelle m. taylor-robinson
To examine how electoral rules and clientelism affect the behavior of elected
officials, I analyze whether deputies initiate bills with a local target.∞≠ These projects,
if passed into law, represent concrete benefits for a community. A school, a road, or
electricity is a project for which the deputy can credibly claim credit, which can
further his political career, or enhance his reputation as a patron.∞∞
Initiating local legislation is only one of many activities for deputies. They have
other legislative activities, such as initiating other types of bills, committee work, and
debates; party duties, such as campaigning and training local party activists; and
from caudillismo to democracy: clientelism in honduras 113
Scott 1969). Desposato (2001) also used this indicator to compare incentives faced by
politicians across Brazilian states to trade votes on executive bills for pork. While
poor people can have preferences about national policy issues, the difficulty of
monitoring and enforcing such policies and the delay before a benefit is received
should cause poor constituents or clients to place a higher value on local public
goods than on national politics. The variable percentage of department population
with no education ranges from 11.6 to 55.1 percent (mean = 32.6, SD = 10).
Clients are expected to be able to demand a higher payment for their vote where
patrons are competing for clients. Party competition is a proxy indicator for competi-
tion among patrons. Beginning in 1993, municipal governments were elected on a
separate ballot, allowing voters to split their vote and giving parties new information
about their support in different localities. Department dominance ranges from 0 for a
department where the two major parties won an equal number of municipal elec-
tions to 50 for a department where one party won all municipal elections (mean =
30.4, SD = 15.5).∞π This measure is available only for the 1994–97 Congress, since
before 1993 municipal elections were part of the fused ballot that elected Congress
and the president.
Models also include a binary variable coded 1 if the deputy is from the president’s
party and 0 otherwise. Number of terms in Congress ranges from 1 to 4 (mean = 1.6,
SD = 0.8), and 1 indicates a deputy’s first term.
findings
Various models assess the impact of electoral and clientelism incentives on the
propensity of deputies to initiate local legislation. Colinearity between district mag-
nitude and percentage of department population with no education necessitates sepa-
rate models (partial correlation of –.72). The strong negative relationship between
DM and poverty measures is not surprising. When a government has very limited
budgets, it is logical to invest resources where the population density is highest, to
produce the greatest popular impact. Citizens often ‘‘vote with their feet,’’ mov-
ing where resources and opportunities are greatest, so the wealthiest departments
have the highest populations. Since representation in the Honduran Congress is
not intentionally malapportioned, the most populous departments have the largest
DM.∞∫ Other measures of electoral incentives (e.g., reelection rates, marginal list
position) can be used in the same model with measures of clientelistic incentives.
Model 1 contains the DM variable as one measure of how electoral incentives
affect the propensity of deputies to initiate local bills, but does not include the
poverty variable (table 5.1).∞Ω The model provides strong support for hypothesis 2—
116 michelle m. taylor-robinson
table 5.1
Impact of Electoral and Clientelistic Incentives on the Propensity of
Honduran Deputies to Initiate Locally Targeted Bills, One Congress Term (1994–97)
Model 1 Model 2 Model 3
table 5.2
The Probability of a Deputy Initiating a Locally Targeted Bill
SIMULATIONS BASED ON MODEL 1
If the deputy is a member of:
the President’s Party the Opposition
If DM of the department is set at the:
Minimum (DM = 1) .80 .26
(.63–.91) (.12–.46)
Mean (DM = 11.7) .58 .11
(.46–.71) (.05–.22)
Maximum (DM = 23) .32 .04
(.17–.51) (.01–.12)
better-off departments where the population has a higher educational level. How-
ever, contrary to expectations in the clientelism literature that patrons will deliver
more services to clients where they face competition for client support, in Honduras
deputies are most likely to initiate locally targeted bills if they are elected from a
department that is dominated by one party. This indicates that when parties and
government face severely constrained resources, they target state resources to reward
and retain existing supporters, not to seek out new clients.
Hypotheses derived from the electoral institutions literature find less support.
DM has the expected effect on deputies’ propensity to initiate locally targeted bills,
but it cannot be tested directly against the poverty variable. Other variables measur-
ing an electoral incentive to initiate local bills did not perform as predicted. Election
from a marginal position on the party’s list does not increase the probability that a
deputy will initiate local legislation. The reelection rate in a department also did not
have a significant effect on the propensity to initiate local bills. Thus it seems that
more than a decade after the installation of democracy, clientelism is influencing
deputy behavior.
table 5.2
Continued
SIMULATIONS BASED ON MODEL 3
If the deputy is a member of:
the President’s Party the Opposition
If percentage of the
department’s population with
no education is set at the:
Minimum (11.6%) .23 .03
(.09–.44) (.00–.09)
Mean (32.5%) .59 .11
(.48–.70) (.04–.23)
Maximum (55.1%) .88 .40
(.72–.97) (.17–.69)
If percentage of the
department’s population
with no education is set at
the:
Minimum (11.6%) .19 .43 .02 .07
(.07–.40) (.15–.73) (.00–.08) (.01–.23)
Mean (32.5%) .53 .77 .09 .22
(.39–.66) (.51–.93) (.03–.20) (.07–.49)
Maximum (55.1%) .86 .94 .34 .61
(.66–.96) (.78–.99) (.13–.64) (.26–.89)
If percentage of the
department’s population
with no education is set at
the:
Minimum (11.6%) .12 .35 .01 .05
(.02–.33) (.15–.60) (.00–.05) (.01–.17)
Mean (32.5%) .37 .73 .05 .18
(.16–.62) (.58–.85) (.01–.16) (.07–.36)
Maximum (55.1%) .75 .93 .22 .54
(.44–.94) (.81–.98) (.05–.55) (.24–.81)
Note: Probabilities were computed holding other variables at their mean; 95% confidence intervals in
parentheses. DM = district magnitude.
120 michelle m. taylor-robinson
Simulations based on model 3 (table 5.2) show that there is an 88 percent proba-
bility that a deputy from the president’s party will initiate a locally targeted bill if she
represents one of the poorest departments. When the educational level of the depart-
ment is comparatively high, there is only a 23 percent probability that a governing
party deputy will initiate a local bill.≤≥ While opposition deputies are overall less
likely than deputies from the governing party to initiate locally targeted bills, they are
most likely to initiate a local bill if they represent a department with an extremely low
level of education. This would seem to indicate that when deputies represent dis-
tricts that fit the conditions where patron-client relations should still be the norm,
politicians are likely to act like patrons who deliver services to their district, using
their position as deputy to enhance their reputation as a patron.
The number of terms the deputy has served in Congress enhances this relation-
ship. A senior deputy from the president’s party who represents a department with
very low educational levels has a 94 percent probability of initiating a local pork bill.
Senior opposition deputies are also quite likely to initiate local pork. There is a 61
percent probability that an opposition deputy with four terms of experience in
Congress representing a department where a high percentage of the population has
no education will initiate a local bill. Politicians who have been in Congress since
the beginning of the democratic regime seem to view delivering pork to their district
as part of their job. Newer deputies are still likely to initiate local bills if they
represent very poor districts, but they are less likely to do so.≤∂ This may indicate a
generational change, or that deputies from poor districts who deliver pork are more
likely to be reelected.
Deputies are more likely to initiate local pork bills in departments dominated by
one party, which is contrary to the expectation that patrons will pay a higher price for
votes in competitive areas. Simulations (not shown in table 5.2) show that when
dominance is at a minimum (i.e., both parties won an equal number of municipal
elections in the department), there is a 16 percent probability that a deputy will
initiate local legislation, but if one party dominates a department, the probability
increases to 46 percent. If the percentage of a department’s population with no
education is set at its maximum value and no party dominates the department, there
is a 75 percent probability that a deputy from the president’s party will initiate a local
pork bill. However, if one party dominates that same department, the probability
that a governing party deputy will initiate local legislation increases to 93 percent
(table 5.2).
The findings of the logistic regressions in models 1 through 3 are robust to an
analysis that examines two terms of legislation (1990–93 and 1994–97), dropping the
department dominance variable that was not available for the 1990–93 term.≤∑ Simu-
from caudillismo to democracy: clientelism in honduras 121
lations based on models with two terms of legislation data (not presented) yield
results similar to those from models 1 through 3. However, the probability that a
deputy from the governing party will initiate a local pork bill is somewhat lower.
When DM is set at the minimum, simulations based on model 1 show that a govern-
ing party deputy has an 80 percent probability of initiating a local bill (table 5.2),
while simulations based on a model using two terms of legislation data predict a 59
percent probability (not shown). When the percentage of a department’s population
with no education is set at the maximum, simulations based on model 3 show that
a governing party deputy has an 88 percent probability of initiating a local bill
(table 5.2), while simulations based on a model using two terms of legislation data
predict a 62 percent probability (not shown). This decrease could be due to the
greater access to the executive branch that deputies from the PNH had during the
presidency of Rafael Callejas (PNH, 1990–94), compared with the access PLH
deputies had to the president, ministers, and agency heads during the presidency of
Carlos Roberto Reina (PLH, 1994–98). PLH deputies complained during interviews
that President Reina did not meet regularly with his party’s caucus in Congress, and
deputies had difficulty getting appointments with ministers and agency directors.
Instead, the executive branch began dealing directly with municipal governments,
and the deputy was no longer an essential facilitator for getting the national govern-
ment to attend to local government and community needs. Only more data from
later administrations will reveal whether this is a permanent change in executive-
legislative relations.
discussion
This chapter examines the incentives that electoral institutions and clientelism
create for legislators to attend to the needs of local communities. Electoral institu-
tions prompt deputies to represent local interests when these institutions create
incentives to seek a personal vote. Clientelism creates different incentives for a
deputy to attend to the needs of communities and individuals. If the deputy wants to
maintain or expand his status as a patron who can be counted on to solve community
problems and to help individuals with their needs, he must deliver services. These
two incentive structures can be reinforcing—a politician who wants to maximize his
personal vote can work to address the needs of clients as part of his personal vote-
seeking strategy. However, even if electoral rules do not give politicians an incentive
to seek a personal vote (e.g., a closed-list PR electoral system with large district
magnitude), legislators can still respond to the incentives created by the informal
institution of clientelism. Deputies who view their role in Congress as acting as a
122 michelle m. taylor-robinson
patron for residents of their district, or who sought a seat in Congress to expand their
resource base for attending to clients, can use the opportunities their position pro-
vides to attend to client needs, even if seeking a personal vote is not necessary for
their reelection. This resource base includes the ability to initiate legislation to target
state resources toward the deputy’s client communities (the behavior examined
here), connections to the executive branch that facilitate lobbying for community
development projects, and helping clients get jobs, scholarships, and licenses.
The analysis finds that deputies are more likely to initiate locally targeted bills
when elected from small-magnitude districts—in other words, when there is some
electoral incentive to seek a personal vote. However, other measures of the incentive
generated by the formal electoral institutions (i.e., marginal list position, reelection
rates in the deputy’s district) did not affect attention to the district. Clientelistic
incentives to attend to local needs, as measured by the poverty of the district, in-
creased the probability that a deputy would initiate locally targeted legislation. This
result held regardless of reelection rates in the district or whether the deputy’s list
position was marginal or safe (i.e., whether electoral incentives were present or
absent). Thus, the incentive for attention to local communities created by clientelism
can help explain behavior that seems odd if only electoral incentives are considered.
In Honduras, electoral incentives for deputies to represent local interests are
weak. The electoral system (closed-list PR with a high average DM and, until 1997,
fused elections for Congress and the president) creates a strong incentive to seek a
partisan vote. Still, some deputies take it upon themselves to initiate locally targeted
legislation, though most do not (84 of the 256 deputies in the dataset initiated local
bills). Many Honduran deputies initiate no bills during a term in Congress (96
deputies in the dataset initiated no bills of any type), which makes the initiatives of
those deputies who did target bills at local communities even more interesting.
Incentives created by clientelism help explain this behavior.
Backbench deputies in Honduras have an incentive to be seen and not heard.
National party leaders control nominations and list position, and they seem to re-
ward quiescent backbenchers. For example, deputies who initiate no national-level
bills have a greater chance of reelection to Congress (Heath and Taylor-Robinson
2003). Yet, some deputies use their time in Congress to target state resources at
specific communities. This is rational behavior for a patron-deputy, even if it is not
necessarily useful for the deputy’s political career. The rationality of this behavior is
enhanced because deputies have some ability to select the type of district they
represent. An aspiring deputy may self-select to represent a small, rural, traditional
department if she wants to play a patron role, while a an aspiring deputy who wants
from caudillismo to democracy: clientelism in honduras 123
to make national policy or represent sectoral interests may seek to run for office in a
large and urban department.
The electoral law allows a deputy to represent the department where she was born
or where she currently resides. Of course, the aspiring deputy still needs to obtain an
‘‘electable’’ place on her party’s list, and national party leaders make those decisions.
But the reality of Honduran politics is that many deputies who represent the depart-
ments that include the country’s major urban areas are not natives of those depart-
ments, while many deputies who represent provincial departments were born in the
department but now live in the capital. Since patronage networks in Honduras work
through political parties, seeking election to Congress to enhance or expand one’s
client network is a risky strategy for patrons, because they will obtain few resources if
their party is in the opposition—only the ability to initiate bills and only a low
probability that those bills will become law. Being a deputy, however, increases a
politician’s or patron’s profile in her community, and if the deputy’s party wins the
presidency, she will be able to take part in the distribution of the spoils of victory.
It makes sense that Honduras’s traditional political parties would support clien-
telistic behavior by their deputies. The electoral system creates little incentive for
elected officials to represent local interests (Shugart and Carey 1992; Taylor 1996).
Lack of attention to local needs is criticized as one of the failings of democracy in
some PR systems in which national party elites control nominations (e.g., Vene-
zuela) (Shugart 2001). In Honduras, however, voters continue to give overwhelming
support to the traditional parties.≤∏ If the traditional parties want to maintain their
popular support base, they need to maintain their links to their clientelistic networks.
This raises an important point about the sources of informal institutional stability
in Honduras. Clientelism has been part of the Honduran cultural landscape for so
long that its persistence is more frequently taken for granted than explained. Any
explanation of the stability of clientelism in Honduras must begin with the fact that
the dominant parties benefit from it. Clientelism is an important asset for the tradi-
tional parties as they work to maintain their political dominance within the con-
straints imposed by free and fair, competitive, democratic elections. To continue to
win control of the executive branch, the Liberal and National parties must obtain
the support of voters. Delivering clientelistic benefits is a cheap way to do so that has
the added benefit of not threatening traditional elite interests. Honduras’s traditional
parties were created as clientelistic networks competing for access to state resources,
not over policy or ideology, and they continue to function in the same manner today.
In the context of an overwhelmingly poor electorate, providing clientelistic ser-
vices is a logical way for parties to compete for voter support. If parties do not provide
124 michelle m. taylor-robinson
the expected level of service to their clients when they win the presidency, party
supporters/clients have shown that they will stay home rather than support the party
in the next election.≤π Hence, it is beneficial to the traditional parties if their back-
benchers use their time in Congress to build their local client base, because those
clients will support their patron’s party in the next election.
In sum, clientelism seems to be a ‘‘substitutive’’ informal institution in Honduras,
providing at least some representation of local interests in the legislature even when
the formal electoral rules do not provide an incentive for elected officials to repre-
sent local needs. In other words, clientelism seems to help Congress perform the
representative function it is charged with in democratic theory.
chapter six
susan c. stokes
Political rules may be written down, or they may not be. And the punishment
triggered when rules are violated may happen in officially sanctioned channels, or it
may not. This is the essence of the distinction between formal and informal institu-
tions that Helmke and Levitsky make in their introduction to this volume. But there
are other dimensions on which rules vary. They may, for instance, be explicitly
understood by those who follow them, or they may be only implicitly understood.
When rules are implicit, even people following them quite closely are frequently
unable to render a formalized account of them. Consider the difference between the
rules of a game and the rules of a language. The rules of chess are explicit; to be able
to play, one has to know the rules. If I do not know, say, that the rook may move an
unlimited number of spaces in horizontal or vertical straight lines, whereas the king
may move only one space but in any direction, I cannot play the game. But if I do
know these rules, I can both make them explicit and play by them.
The rules governing the speaking of a language are quite different. We learn to
follow these rules well before we learn what the rules are—if indeed we ever learn
what they are. A quite average five-year-old will utter many sentences of his or her
‘‘native’’ (first, early-acquired) language in ways that conform to the formalized
grammatical rules of that language, whereas even a highly intelligent adult speaker
of the same language will have great difficulty explaining these rules unless trained
to do so.∞
126 susan c. stokes
I call rules that can be rendered explicit by those following them game rules. I call
rules that are implicit and difficult to formalize by those using them grammati-
cal rules.
Two central premises of this chapter are (1) that the way people vote in elections is
rule-governed—that the phrase voters’ decision-rules is to be taken seriously—and (2)
that voters’ decision-rules are grammatical. They are rules, and not simple responses
to physiological or material incentives; they are diffused through social mechanisms,
not merely thought up by individuals; they are not normally rendered explicit, at
least not in a formalized version, by people who use them; and, to varying degrees
depending on the rule, they are violated at the risk of social—and even, sometimes,
physical and material—sanction.
Much is at stake for democracies in the informal and grammatical decision-rules
that voters adopt. Consider eight rules, each of which has a large political science
literature behind it, claiming that it is the rule that voters use.
These decision-rules, when adopted by large numbers of voters, have major and
varying implications for how well democracy works. At least in theory, retrospective
voting can induce accountability if it is sociotropic, but not if it is egocentric (Fere-
john 1986). Spatial, strategic, and directional decision-rules hold out some hope for
accounting for accountability in argentina 127
prop up that country’s formal institutions of democracy when the latter really col-
lapse. Rather, my claim is that democracy functions better there (and probably in
most democracies, new and old) when informal rules enforce accountability.
I begin by noting substantial differences in the quality of democracy across towns,
cities, and provinces in Argentina. I then pose the question, do these differences
correspond to systematic differences in the electoral decision-rules that voters em-
ploy across regions? Are there distinct dialects, if you will, in the grammatical rules
underlying voters’ choices? I draw on survey research indicating that the answer is
yes. People in regions where democracy works relatively well display a heightened
appreciation for mechanisms of accountability and, more than people from other
regions, expect others—their neighbors and officeholders—to follow decision-rules
that make accountability possible. People in regions where democracy has worked
less well expect their neighbors to follow voting rules that are supportive of clientel-
ism. I do not offer here a full account of the origin of these expectations and rules,
but I do eliminate some explanations that the literature on the transition and consoli-
dation of democracy suggests, and speculate about a better answer. Neither the
income levels of individuals nor social capital explain regional differences in the
prominence of rules of accountability. What may matter more are traits not of
individuals but of communities: their levels of social equality. I conclude by return-
ing to the general question of the importance of informal institutions to democracy.
Local democracy works better in some parts of Argentina than in others.∏ Some
provinces are good at maintaining balanced budgets; others are profligate spenders.
Some provinces strive to implement good public policy; in others, public assistance
is glossed as personal charity by the governor. Experts deem the Argentine federal
budget and the process by which it is written to be far more opaque than in other
Latin American countries; yet a handful of mayors have tried to bring community
organizations and individuals into the process, at least at the local level. In some
cities, powerful political and economic actors impose their will on the citizenry; in
others, the political leadership has experimented with relatively novel forms, such as
the referendum, to give residents a voice in crucial collective decisions. These
comparisons allow Cleary and Stokes (forthcoming) to posit a ranking for a set of
four regions in which we conducted sample surveys. Among these four regions, we
ranked the city of Mar del Plata and the surrounding district of General Pueyrredón
the highest on measures of democracy, followed by the rest of the provinces of
130 susan c. stokes
table 6.1
Responses to Questions about Institutional Trust and
Trust in Politicians by Region, Argentina
Mar del Buenos
Plata Aires Córdoba Misiones Total
Services: When governments provide good services to the people, is this because
—they are under the watch of the courts, 65% 56% 48% 40% 52%
Congress, or the press? (311) (268) (232) (192) (1,003)
—they are good, committed people? 30% 40% 40% 53% 41%
(142) (192) (194) (256) (784)
No answer 6% 4% 11% 7% 7%
(27) (20) (54) (32) (133)
Efficient: When governments function efficiently, is this because
—they know if they don’t, people won’t 71% 75% 68% 67% 70%
vote for them in the next election? (340) (362) (326) (321) (1,349)
—the people governing are good, 24% 22% 26% 29% 25%
committed people? (116) (106) (125) (140) (487)
No answer 5% 3% 6% 4% 4%
(24) (12) (29) (19) (84)
Attention: When politicians really pay attention to people like you, is this because
—they want to be reelected? 85% 80% 78% 78% 81%
(410) (386) (375) (375) (1,546)
—they really care? 11% 17% 16% 18% 16%
(55) (83) (76) (87) (301)
No answer 3% 2% 6% 4% 4%
(15) (11) (29) (18) (73)
Buenos Aires, Córdoba, and, finally, the province of Misiones, where case studies of
local politics find an extensive clientelism, personalism, and extreme domination of
political life by powerful economic actors.π
Do people in Mar del Plata perceive the motivations for politicians’ actions and
the constraints they are under differently than people in, say, Misiones? In the three
survey questions listed in table 6.1, we asserted that politicians sometimes behave
well—provide good public services (services), are efficient (efficient ), pay attention to
the opinions of constituents like the respondent (attention)—and then asked re-
spondents to choose among several alternative explanations for this good behavior.
In each case one option attributed good behavior to the personal qualities of politi-
cians (‘‘they are committed people,’’ ‘‘they really care about constituents’ opinions’’),
whereas the other attributed it to mechanisms of accountability (‘‘they are under the
watch of the courts,’’ ‘‘people won’t vote for them in the next election’’). Among
respondents across all three provinces, the ‘‘accountability’’ answer was in all cases
the one respondents chose more frequently.
Responses to these questions differed systematically by region. The number of
respondents offering ‘‘accountability’’ responses to the questions in table 6.1 is high-
est in the most democratic region and then trends downward. For example, 65 per-
accounting for accountability in argentina 131
cent of respondents from Mar del Plata attributed good services to institutional
monitoring, while only 40 percent of respondents from Misiones made such an
attribution. In many cases, institutional trust declines monotonically as the level of
democracy declines.
To probe systematic differences across regions, we coded answers to services,
efficient, and attention as dummy variables for an ‘‘accountability’’ answer. Thus for
a person who chose the answer ‘‘when municipal governments are efficient, this is
because otherwise people won’t vote for them in the next election,’’ the score is 1 on
the dummy variable efficient; otherwise the score is 0. We then estimated logit
regression models of these accountability answers. The results (table 6.2) show that
older people were more skeptical of the characters of politicians and more likely to
ascribe good behavior to mechanisms of accountability. Women were more likely to
ascribe good government to the characters of politicians; men, to politicians’ fear
of losing votes. Most important for our purposes, respondents in Buenos Aires,
Córdoba, and Misiones were generally less likely to offer ‘‘accountability’’ responses
than were respondents in Mar del Plata.∫ They were significantly less likely than
were people from Mar del Plata to say that governments provide services when they
are under the watch of the courts, the congress, and the press; somewhat less likely to
say that efficient governments are ones that fear losing office; and significantly more
likely to say that politicians who pay attention want to be reelected.
The regional effect on expectations of accountability was sometimes quite strong.
To give a sense of how much of a difference region makes in institutional trust,
consider a typical respondent in our sample, one with an average household income,
educational level, and quality of housing, and living in a city of average size. Simula-
tions show that if she lived in Mar del Plata, this typical respondent had a 68 percent
chance of saying that governments provide good services ‘‘because they’re moni-
tored.’’ But if she lived in Misiones, the chance dropped to 41 percent.Ω
Expectations of accountability, when they do occur in Argentina, focus both on
formal institutions and on informal rules and behaviors. In some settings people
believed that politicians performed well when under the watchful eye of the courts or
other branches of government. But they also believed in informal rules of account-
ability. In more democratic regions, the belief was more widespread that anticipation
of the ‘‘future retrospective judgment of voters’’ (Manin 1997) or anticipation of ‘‘the
moment when their power is to cease, when their exercise of it is to be reviewed’’
(Madison et al. 2000 [1788]) could induce governments to be efficient. The electoral
connection involves several informal rules and the expectation that other people will
follow these rules. For the connection to hold, voters must expect other voters to use
appropriate criteria in deciding how to vote. Consider again the question ‘‘when
132 susan c. stokes
table 6.2
Logit Models of ‘‘Accountability’’ Responses to Questions about Politicians
Model 1 Model 2 Model 3
Services Efficient Attention
municipal governments work efficiently, is this because they are staffed by good, com-
mitted people, or because they know that if they don’t work well people won’t vote for
them later?’’ To answer ‘‘because they know that if they don’t work well people won’t
vote for them later,’’ respondents have to believe that other voters do consider the
incumbent government’s efficiency when they decide how to vote. They must expect
other voters to use retrospective criteria and not to be bought off by the small, individ-
ualized inducements of clientelism, which tend to come at the cost of efficiency and
the provision of public goods (Estévez et al. 2003; Medina and Stokes 2003).
accounting for accountability in argentina 133
table 6.3
Responses to Questions about Voters’ Decision Rules
Non-
Accountability accountability No
Question Variable response response answer
The survey provides more direct evidence that the expectation that one’s neigh-
bors would follow retrospective and not clientelistic decision-rules in deciding how
to vote was more widespread in Argentina’s higher-performing democratic regions
than in its lower-performing regions. We asked survey respondents questions about
how they vote and how they expected other people in their neighborhood to vote
(table 6.3). Table 6.4 reports logit regression estimations. In the first two models, the
dependent variables are dummies for people who said that their neighbors used
clientelistic decision-rules, rules that would not support accountability: that they
responded to handouts (handout ) or favors ( favor), not a party’s program or its
concern for everyone, when deciding how to vote. The negative coefficient on
dummies for Mar del Plata (Mar del Plata) indicates that respondents from that city
perceived their neighbors as less clientelistic than did respondents from other cities
and towns. People from Mar del Plata were also less likely to be the targets of
clientelistic mobilization. The dependent variable in the third model (gift ) is a
dummy for people who reported that they had received a handout from a party
operative in the last campaign. The negative sign on the Mar del Plata coefficient
shows that people there were less likely to have received such handouts, leaving
them freer, presumably, to take performance into account.
In one way, Mar del Plata residents were less prone to accountability than people
from elsewhere. We asked whether people in the respondent’s neighborhood paid
attention to how well a party performed in the past versus how good its proposal
was for the future (past ). People in Mar del Plata viewed their neighbors as more
prospective than did people elsewhere (as indicated by the negative sign on the
134 susan c. stokes
table 6.4
Logit Models of Clientelistic (Non-Accountability) Responses to Questions about Voting
Model 1 Model 2 Model 3 Model 4
Handout Favor Gift Past
coefficient relating the Mar del Plata dummy to past in model 4; table 6.4). As
Fearon (1999) explains, retrospective and prospective considerations can interfere
with one another when voters attempt to use their vote to induce politicians to be
responsive.
Yet the central finding thus far is that people who live in regions where de-
mocracy functions fairly well, in a national context where democracy more often
performs badly, differ from people from other cities and towns in that they are
more likely to abide by informal rules of politics that support accountability, and
they are more likely to expect their neighbors also to abide by these rules. They
also show a stronger appreciation of formal institutions that support democratic
accountability.
accounting for accountability in argentina 135
Social scientists have emphasized in recent years the importance of trust, and of
the institutions that promote trust, in making social relations work smoothly, both
market relations (see esp. North 1990) and relations between citizens and representa-
tives (Putnam 1993; Levi 1997; Ferejohn 1999). Are people from more democratic
regions of Argentina particularly trusting?
We asked a series of questions designed to measure levels of interpersonal trust.
By trust I mean A’s belief that B will act in A’s interest, even though B would stand to
gain in some way by not so acting and even though A cannot directly monitor B’s
relevant actions (Cleary and Stokes, forthcoming). Table 6.5 lists questions that
explore levels of interpersonal trust: whether respondents trust their neighbors, and
whether they view other people in the abstract as trustworthy. Respondents across the
three provinces were quite trusting of their neighbors, saying by a margin of three to
one that they would trust a neighbor to care for their home while they were away. But
they displayed widespread distrust of abstract others, with majorities agreeing that
most people will take advantage of you when they can, and that a minority of people
are trustworthy. Several studies indicate that Argentines came to trust each other less
and less over the first two decades of renewed democracy. The percentage respond-
ing in the World Values Survey that one could trust other people fell from 26 percent
in 1984 to 18 percent in 1995 (Mussetta 2002, 66).
If democracy worked unusually well in some regions of Argentina, this was not
because people there were more trusting in a generalized sense than were people in
other regions. Consider the question ‘‘thinking about the locality or barrio where you
live, if you were to go on a trip, do you have any neighbors whom you could trust to
care for your house while you were away?’’ (variable neighbor). To discern the factors
influencing people’s answers, we estimated logit regression models (not shown) in
which the dependent variable was a dummy that took the value of 1 when a person
said she would leave her house in the care of a neighbor, 0 when she said she would
not. Information about individuals’ income and the quality of their housing, and
about the size of the city or town they live in, would help us predict their answer to
this question. (People with relatively high incomes, who lived in high-quality hous-
ing, and who were from smaller towns and cities were more likely to trust their
neighbors.) But knowing the region where the person lived would not help us predict
the answer.
Expectations of accountability among residents of regions where democracy
136 susan c. stokes
table 6.5
Responses to Questions about Personal Trust, by Region, Argentina
Mar del Buenos
Plata Aires Córdoba Misiones Total
Neighbor: If you go away on a trip, do you have a neighbor whom you could trust to care for your
house?
Yes 75% 80% 75% 68% 75%
(360) (383) (359) (328) (1,430)
No 25% 20% 24% 30% 25%
(113) (97) (117) (146) (473)
No answer 2% 0% 1% 1% 1%
(7) (0) (4) (6) (17)
Advantage: Do you believe most people would take advantage of you if they had the chance?
Yes 55% 53% 50% 57% 54%
(263) (254) (241) (272) (1,034)
No 42% 45% 45% 37% 42%
(203) (215) (214) (178) (810)
No answer 3% 2% 5% 6% 4%
(14) (11) (25) (30) (80)
Trust: Which is closest to your way of thinking?
You can trust a majority of people. 20% 26% 23% 18% 22%
(96) (123) (111) (84) (414)
You can only trust a minority. 61% 61% 56% 59% 59%
(294) (291) (271) (284) (1,140)
I don’t trust anyone. 18% 13% 20% 22% 18%
(87) (64) (95) (105) (351)
No answer 1% 0.4% 1% 2% 1%
(3) (2) (3) (7) (15)
worked unusually well focused on politics, and did not extend to commercial rela-
tions. Our survey asked: ‘‘Considering now merchants who act honestly, is this
because (1) they are honest people, (2) if they weren’t, they know they could be fined,
or (3) they could lose clients?’’ Among all respondents, 45 percent answered ‘‘be-
cause they are honest’’; 8 percent, ‘‘because they know they could be fined,’’ and 45
percent, ‘‘because they could lose clients.’’ One can think of people who answered
‘‘they know they could be fined’’ as believing in formal, legal institutions of commer-
cial accountability, and those who answered ‘‘they could lose clients’’ as believing in
informal rules—there is no law that says customers must abandon merchants whom
they discover to be dishonest, and no sanction for failing to abandon them, but we
would expect many customers to do just this. Such beliefs (as well as formal legal
institutions) help reduce transaction costs and smooth the functioning of market
economies (North 1990). But what is important for our purposes is that residents of
Buenos Aires (including Mar del Plata) were significantly less likely to believe in
either a formal or an informal institution of market accountability than were people
in Córdoba and Misiones.
accounting for accountability in argentina 137
The question that immediately suggests itself is, do these informal (and formal)
rules, and the expectations that others will follow them, in fact cause democracy to
function better? Or, instead, is the causal sequence reversed, and relatively lively
local democracy causes beliefs in these informal rules? Why does democracy func-
tion relatively well in some settings, and are informal rules of accountability really
the cause? Although I will not attempt here a full answer to these questions in the
Argentine case, I offer some preliminary thoughts.
Notice first some explanations that we can reject. One line of thinking, proposed
most prominently by Robert Putnam (1993), focuses on a political culture of trust.
Some regions in a country may develop a relatively rich associational life, a greater
reservoir of personal trust, and hence larger supplies of social capital. According to
Putnam, democracy works better in the north than in the south of Italy because ‘‘the
civic community’’ in the north ‘‘is marked by an active, public-spirited citizenry, by
egalitarian political relations, by a social fabric of trust and cooperation,’’ whereas the
south is ‘‘cursed with vertically structured politics, a social life of fragmentation and
isolation, and a culture of distrust ’’ (1993, 15; emphasis mine). Social trust in the
polity acts like market trust in the economy, allowing people to achieve higher levels
of efficiency. A trusting society secretes a well-functioning democracy.
In Argentina, as we have seen, people in places where democracy worked rela-
tively well did not trust people in general, or their neighbors, or merchants, or
politicians, more than did people in regions where democracy worked less well.
They were not particularly prone to entrust their homes to their neighbors, but they
did trust their fellow voters to punish politicians who underperformed, and they
trusted their fellow voters not to be bought off by handouts. And they did not trust
politicians except in the convoluted sense of trusting them to follow their own
interest in staying in office and out of jail. A final piece of evidence that social capital
is not the key to our story is that people in more democratic regions do not seem to
have created more social capital than people elsewhere. We asked a range of ques-
tions about respondents’ involvement in organizations and about whether and how
frequently they attended meetings. These measures of social capital failed to predict
138 susan c. stokes
adherence to rules of accountability; nor were people from more democratic regions
more involved in associational life than people from less democratic ones.
Another venerable line of research links the emergence and consolidation of
democracy with economic development. Higher incomes are supposed to imbue
people with longer time horizons and make them less conflict-prone and more
tolerant of others’ opinions—all of which are in turn supposed to promote democ-
racy (see esp. Lipset 1958, 1960). Among people who answered our surveys, we
observed no straightforward effect of income, education, or quality of housing on
beliefs that might promote democracy. Just as important, whatever effects people’s
household incomes or educational levels or quality of housing might have had, they
did not reduce the effect of region on expectations of accountability. If we compare a
resident of Mar del Plata with a resident of a Misiones city or town of the same size as
Mar del Plata, both with the same level of income, we find the Mar del Plata resident
is considerably more likely to abide by, and expect others to abide by, informal rules
that support accountability.
In contrast, income and development, not as traits of individuals but as structural
traits of communities, may help explain differences in the quality of local democracy
in Argentina. Regions and cities vary in their levels of development and in their class
structures. Scholars in comparative politics have contended that equality promotes
democracy (see esp. Boix 2003; Acemoglu and Robinson 2005; this idea also appears
in places in Lipset 1958). When the gap between rich and poor is relatively narrow,
the rich may view the poor as less ‘‘beyond the pale’’ (Lipset 1958, 83) and hence
more readily incorporated into civic life; and social equality may make the poor look
less threatening to the wealthy, because they are less prone to pursue strongly re-
distributive measures (Boix 2003). Applied to our context, one could imagine that a
community that is not socially polarized would also be one in which people would
be relatively ready to expect others to follow courses of action appropriate to sustain-
ing democratic accountability. It is tantalizing to note, along these lines, that a
higher percentage of self-reported incomes in Mar del Plata than in any of our other
three regions were concentrated in the middle of the range (300–700 pesos per
month; 44% reported this as their household income in Mar del Plata, whereas only
33% did in Misiones, with the other two sampling regions falling in between). It is
also tantalizing to note that when asked to place themselves in the lower, middle, or
upper class, a larger percentage of people in Mar del Plata chose ‘‘middle class’’ than
in any of the other regions. Yet these comparisons could be misleading: they in-
volve comparing income distribution in a single city with income distribution across
whole provinces.∞≠
My hunch is that beliefs in accountability are both a cause and a consequence of
accounting for accountability in argentina 139
well-functioning democracy, and that both structural factors and good luck are often
at play. Imagine that a city happens, for no structural reason, to generate a set of
judges and lawyers who aggressively prosecute corruption and waste in city govern-
ment. Citizens who observed politicians being held accountable should display a
growing belief that politicians can be held accountable. Because they hold this be-
lief, they are gradually willing to concede more power and resources to city govern-
ment. Local government does more for people, and its own structures are strength-
ened, including those that strengthen accountability. A virtuous cycle is underway.
At the same time, structural factors, such as development and equality, are likely to
further strengthen both the reality of, and expectations of, local democracy.
conclusion
When democracy works well, when it achieves the effects that make it better than
other systems of government, this is because formal institutions, informal institu-
tions, and informal rules of behavior interact in felicitous ways. The new institu-
tionalism in comparative politics may have underestimated the importance of infor-
mal institutions and informal rules—whether they are more like the rules of a game
or the rules of grammar—for democratic outcomes. Comparativists have tended to
conceive of institutions as creating incentives that map straightforwardly onto out-
comes. A better way to think about the mapping of formal institutions on outcomes is
that their effect is often conditional on informal institutions and rules, and these
informal institutions and rules are variables that need to be unpacked empirically. As
Helmke and Levitsky note in their introduction, the role of informal rules in democ-
racy goes beyond one of filling the gaps left by formal institutions. In fact, formal
institutions in some instances will not work in the way they were meant to unless the
appropriate informal rules are in place.
We have seen a lot of evidence from Argentina that certain informal rules covary,
at least, with good democratic outcomes. What seems to have mattered for a rela-
tively lively local democracy is not that people trust politicians or other people in the
simple sense of expecting them to promote one’s interests if unconstrained. Expecta-
tions of accountability, rather than trust, were the informal rules that helped support
local democracy in Argentina.
This page intentionally left blank
part iii
joy langston
him as the PRI’s presidential candidate, and to impose that choice on the party’s
political elite. The PRI candidate would then go on to win a noncompetitive general
election. Related informal rules later grew up to support this informal prerogative,
including a restriction of the universe of possible successors to the president’s cabinet
and a prohibition against openly seeking the nomination or criticizing other poten-
tial nominees. These rules of the game were institutionalized during the 1940s and
1950s and remained intact into the mid-1990s.
The dedazo resolved some important dilemmas for the post-revolutionary elite.
Neither formal party statutes nor the competitive elections prescribed by the Consti-
tution could achieve this elite’s two main goals: maintaining PRI dominance while
avoiding an internal rupture. By ensuring regular leadership turnover while at the
same time creating mechanisms to limit internal conflict, the dedazo helped put an
end to the military rebellions and political defections that had threatened the regime
since the 1920s. Once institutionalized, informal rules drove up the costs of disobey-
ing regime leaders and made disloyalty by a member of the PRI a losing proposition.
Thus, party politicians chose to remain within the governing coalition, helping to
make the PRI practically invincible for seven decades.
In this chapter, I examine the origins and evolution of the dedazo. Whereas much
research has gone into questions of formal institutional creation and change, little
has been written on the origins, evolution, and collapse of informal rules. The
chapter shows how, beginning in the administration of Lázaro Cárdenas (1934–40),
PRI elites devised a set of informal rules and procedures that minimized the possi-
bilities of rupture. Regime leaders used a series of instruments available to them
because of their hegemonic control over the state, including electoral fraud, changes
to the electoral laws, and a variety of selective political benefits, to reward PRIistas
who played by the rules and punish those who broke them. The process of learning
and adapting to the informal rules of presidential succession was slow and evolu-
tionary. But once top leaders accepted the rules and it became clear that under-
lings lacked the power to contest them, a stable equilibrium emerged that endured
for decades.
I also examine the collapse of the dedazo. Like formal institutional change,
informal institutional change occurs for a variety of reasons. In Mexico, important
changes in the external environment—particularly the rise of electoral competition,
which created opportunities for politicians inside the PRI—led to a change in formal
rules, which in turn allowed ambitious party politicians to dismantle the underlying
informal power arrangements. After the onset of electoral competition, the condi-
tions were set for a new bargaining game between presidents and ambitious PRI
politicians. Presidents had power over ballot access and key financial and patronage
transformation of the dedazo in mexico 145
resources, but emerging subnational politicians possessed a resource that was critical
in a context of competitive elections: public support. These politicians could use the
now credible threat to leave the PRI and run under another party label to leverage
concessions from the PRI leadership. The resulting bargaining process produced
both formal and informal rules change, as ambitious politicians used party statutes to
weaken the informal prerogatives that PRI leaders had previously enjoyed in impos-
ing candidacies.
The Mexican story goes against much of what we have learned about the relation-
ship between formal and informal institutions. First, contra distinctions between
formal and informal rules that center on external (state) versus self-enforcement, the
dedazo was a state-enforced informal institution. Although it maintained the fiction
that the Constitution and the PRI’s statutory rules determined electoral outcomes,
the Mexican state systematically enforced these informal arrangements. The conse-
quences for breaking the informal rules were clear and known: government officials
would block political advancement, engineer attacks in the state-controlled press,
or manipulate the judicial system to take legal action, in some cases leading to
imprisonment. Although this enforcement took place outside officially sanctioned
channels, it was clearly backed by the power of the state. Second, whereas informal
institutions are frequently viewed as endogenous to formal institutional arrange-
ments, in the sense that actors create them in response to certain formal rules, the
Mexican case suggests that the relationship may also be reversed. The PRI’s formal
statutes were written (and, in some cases, rewritten) to support the informal rules of
the game. For example, statutory rules were couched in a highly ambiguous manner
so as to lower the costs of utilizing the informal rules. Finally, whereas most of the
existing literature on informal institutions portrays them as changing very slowly, the
dedazo—though it took decades to become fully operational—collapsed relatively
quickly.
could be rewarded in the medium and long run. This informal prerogative to select
lower-level candidates reinforced each president’s ability to maintain control over
the enormous PRI political elite, which was often divided into different ideological
wings and personalized factions.
When and how was the dedazo born? The origins of the dedazo lie in the reac-
tions of coalition leaders to the harsh reality of political anarchy prevalent in Mexico
during and after the Revolution. Post-revolutionary elites faced an important collec-
tive dilemma. If they all cooperated and none rebelled against or challenged the new
regime, then they all would be better off, because a unified governing party would be
able to exclude other groups from power. However, some actors (i.e., those who won
the presidency) benefited far more than others, so there were strong incentives to
attempt an exit and capture of the presidency either by military rebellion or, in later
decades, by competing against the PRI incumbents in the general election. In other
words, competition for the presidency generated severe internal conflicts that threat-
ened to divide the governing party and possibly undermine its grip on power.
Between the end of the Revolution (roughly 1917) and the first clear-cut dedazo
of 1939–40, presidential succession was anything but institutionalized. Presidents
emerged out of armed conflict, elite negotiation, and the imposition of nonelected
strongmen. The initial post-revolutionary period was characterized by instability and
violent conflict. Assassinations and armed rebellions marked the turnover of power
from General Venustiano Carranza’s defeat of Francisco Villa in 1915, to Carranza’s
assassination in 1920, to Adolfo De la Huerta’s rebellion in 1923-24 to protest the
imposition of Plutarco Elías Calles as presidential candidate (Meyer 1985). This
continued in the failed revolt of 1927 (and the shooting of approximately fifty gen-
erals in its aftermath), to General José Gonzalo Escobar’s challenge to Calles’s
authority in 1929. Alan Knight writes that these episodes were ‘‘all examples of a
Darwinian (or Hobbesian) competition within the revolution itself ’’ (1992, 131).
There were no clear rules of succession aside from the ‘‘survival of the fittest’’ and
most savage, which played out in an unending series of violent conflicts to decide the
right to govern Mexico.
During the 1920s, two leading generals, Álvaro Obregón and Plutarco Elías Calles,
worked out a power-sharing plan in which they alternated in the presidency. When
Obregón was assassinated in 1928, Calles became the jefe máximo of the entire
nation. During a short period between 1928 and 1934 (known as the Maximato),
General Calles was able to place his handpicked allies as presidents of the nation
148 joy langston
and, in one case, force the president’s ouster. Most politicians considered Calles the
true leader of the nation, and would often travel to his compound, close to Mexico
City, before conducting business with the elected president. Calles imposed a for-
mer governor of the state of Michoacán and former cabinet minister, Lázaro Cár-
denas, as the presidential candidate in 1934 and expected he would be able to control
this president as he had the last three. However, Cárdenas used his executive power
to exile the strongman Calles and to remove the former president’s allies from
government and elective positions (González Casanova 1965; Lerner de Sheinbaum
and Ralsky 1976; Medina 1977; A. Knight 1990).
The struggles within the governing coalition threatened the very foundation of
the fledgling regime. If the post-revolutionary elite continued to split apart during
each presidential turnover, they would be unlikely to consolidate power. One possi-
ble mechanism of succession was, of course, that prescribed by the Constitution:
competitive elections. Yet such a system would threaten the revolutionary elite’s
hegemonic grip on power. Another option would have been to permit presidential
reelection and thus follow the Soviet or Taiwanese solution of allowing a strong
leader to stay in office until his death. Yet Mexico’s recent experience with the thirty-
year dictatorship of Porfirio Diaz (in which political mobility was blocked and
succession was ultimately achieved only through large-scale violence) created an
incentive for post-revolutionary leaders not to allow executives to legally succeed
themselves, and reelection was prohibited in the 1917 Constitution. Although this
clause was not as well institutionalized as is often believed,≥ the fact that several
revolutionary generals with access to arms were contending for the presidency sug-
gested that reelection would invite violent turnovers in power.
Cárdenas’s solution was to impose his favored successor on the political class and
the nation at large, while not retaining executive authority for himself. When Cár-
denas named Manuel Ávila Camacho to succeed him, a tradition of presidents
imposing their successors and then turning over power to them was born. The
regime’s leaders decided that the best way to avoid overt conflict was for the sitting
president to decide the succession (and allow future presidents the right to do the
same). The new dedazo rule would eventually lower levels of conflict over the
succession, because only one man—the current president—would decide the issue.
This also made the president the ground zero of loyalty until very late in his single
six-year term, because other politicians’ political futures depended on him up until
he had made his decision.
Why did PRI elites choose not to enshrine the dedazo in the Constitution or the
party statutes? There are several reasons why they opted for an informal institutional
solution. For one, they sought to uphold the image that Mexico was an emerging
transformation of the dedazo in mexico 149
democracy, and the formally democratic 1917 Constitution was far better suited to
this purpose than was the authoritarian dedazo. Informality also helped the PRI
leadership maintain a greater degree of flexibility and discretion vis-à-vis opposition
parties and members of the PRI coalition. It was less costly to create and enforce
informal institutions of intraparty governance than to work through formal rules and
procedures. If the openly authoritarian practices had been written down in the party
statutes, they would have been exposed to public scrutiny and debate, and PRI
politicians could have focused on changing them. Unwritten rules were a much
more difficult and ambiguous target.
(Reyna 1985, 108). This move eliminated all PRI governors, former cabinet ministers
(and therefore allies of former presidents), and all federal legislators from consider-
ation. Now those who could aspire to win the nomination had to belong to the
president’s closest circle and thus depend on the good will of the current leader. The
president literally controlled the universe of potential candidates, because he placed
and removed them from cabinet positions and, in doing so, could manage the
conflict over the succession, which now took place within the executive bureau-
cracy. This, the second informal rule, also made cabinet appointments and shuffles
hugely important.∑
A third informal rule that formed the base of the succession process was the
prohibition against openly stating one’s presidential ambitions or too actively engag-
ing in political mobilization within the regime’s highest reaches before the president
had made his decision and communicated it (Reyna 1985). In the 1940, 1946, and
1952 ruptures, groups within the regime had formed openly to support both the
official nominee and the challenger before and after the sitting president had made
his choice known (Garrido 1982, 279, 287; Meyer 1985, 93; Paoli Bolio 1985, 142).
This open mobilization within the regime’s confines had to be stopped, or at least
dampened, if the sitting president were to control his succession process. Thus, an
informal rule was created to block ‘‘pre-candidate’’ mobilization within the regime
and executive bureaucracy. This made it more difficult to openly promote one’s own
candidacy, which had in earlier successions created ‘‘momentum’’ for potential
candidates (both within and outside the PRI) and thus created problems for the
president when he attempted to impose his favorite. As Fidel Velasquez (former
long-lived leader of the PRI’s most important workers’ peak-level association, the
Mexican Workers Confederation [CTM]) once stated: ‘‘He who moves won’t come
out in the picture,’’ meaning that if one mobilizes political support openly, this will
be held against him and he will not be chosen.
In each rupture attempt, the learning process of both regime members and
leaders advanced toward the equilibrium of regime stability and discipline in the
face of nomination defeat. In the early years of the PRI regime, from roughly 1934 to
1958, although none of the regime’s leaders may have thought that the specific
changes or adjustments they made to the succession process would eventually end
up creating a regime that would remain in power for more than seventy years, the
sum of their actions did help lead to this outcome. The historical record leads one
to conclude that Cárdenas (1934–40), Manuel Ávila Camacho (1940–46), Miguel
Alemán (1946–52), and Adolfo Ruíz Cortines (1952–58) all aimed to ease the con-
flicts inherent in the turnover of executive power, and, on taking office, each incom-
transformation of the dedazo in mexico 151
ing president during this period made further incremental changes that also helped
create a less troubled road to the following succession. In all three ruptured succes-
sions, the regime’s leaders used a mixture of positive incentives and credible threats
to keep those supporters of the challengers who were poised to leave firmly within
the ranks of the coalition. As early as the 1940s, the Mexican state had control of sub-
stantial amounts of resources, graft, and public posts (including lower-level elected
positions) that could be selectively doled out. Once politicians had left the party to
back a challenger’s bid, and had seen him lose at the ballot box, they were welcomed
back into the regime and allowed to remake their careers. Simply put, after the
failure of three successive bids (by leaving the coalition) to break the president’s
informal prerogative to choose his successor, PRI politicians did not believe an
electoral challenge would succeed, especially after the 1946 electoral reforms that
centralized the organization, management, and outcomes of the elections in the
hands of the secretary of gobernación.
Another important rule of this period was that no seated bureaucrat or elected
official who had held a position in the six months before the election could run for
the highest office of the land. This forced ambitious presidential hopefuls and their
supporters to make a crucial decision long before they could gauge whether a split
from the hegemonic party would be successful, and gave the regime’s leaders more
time to immobilize the challenger. By the 1952 rupture, there were indications that
an exit would not be successful and, as a consequence, supporters of General Miguel
Henriqúez were determined to win the official nomination rather than leave the
party for an external run at the presidency (Rodríguez Araujo 1974, 108). Because
most politicians, even if they were not close allies of the next presidential candidate,
would eventually find a post within the regime, the potential gains to be had from
remaining disciplined were far higher than the potential gains of a most likely
unsuccessful bid to unseat the hegemonic party at the ballot box. There were few
incentives to challenge the president’s informal right to designate his successor, and
no important PRI politician would do so until the 1987 rupture, which took place
after several years of economic crisis and a profound shift in economic development
model (Garrido 1993; Bruhn 1997).
Regime leaders also created new formal electoral rules to dampen the temptation
to leave the coalition during the presidential succession process. Molinar (1991) has
shown that the 1946 electoral reforms to centralize the electoral process under the
aegis of the secretary of gobernación were an important step in corralling losers
within the coalition, because they substantially raised the costs of forming a new
party. Because party registration was now controlled by Gobernación, it became far
152 joy langston
more difficult for politicians to leave the PRI and run under another party’s banner.
Thus, both formal and informal rules were created to permit each outgoing presi-
dent to single-handedly select his successor.
The dedazo was highly resistant to change; from 1952 until 1987, there were no
serious challenges to the president’s prerogative of choosing his successor. It took a
combined punch of both a disastrous economic crisis (1980s) and growing electoral
competition (1990s) to finally allow ambitious politicians within the party to rework
the tradition of the dedazo. In this section, I examine changes to the dedazo in the
late 1990s that were due to a split in 1987 and to rising levels of electoral competition
and shocks at the ballot box, changes that encouraged PRI members and activists to
use the formal statutory rules to strengthen their position against the informal presi-
dential imposition of his successor.
Informal rules are usually understood to be an instrument for playing an end run
around inefficient or undesirable formal institutions: in the former case—following
Helmke and Levistky’s terminology in the introduction to this volume—informal
institutions are substitutive; in the latter, they are competing. Yet, because the infor-
mal rules played such an important role in political outcomes in Mexico and were
undergirded by executive power and noncompetitive electoral conditions, party
actors who were dissatisfied with arbitrary impositions in the presidential succession
could do little except wait for a future opening. This began to change with electoral
competition. PRI politicians were empowered by electoral competition, because it
gave them an exit option that had been nonexistent under hegemonic conditions.
PRI politicians would end up using the formal party rules to shore up their weak
position and force the president to allow more participation in the final nomination
decision. So, while most scholars of Mexico have ignored the interplay between
informal and formal institutions, believing that only the former matter, the formal
statutory rules of the game in fact empowered weaker party actors to dilute the
informal prerogatives of party leaders.
The 1987–88 presidential succession was undermined by a rupture caused by a
small number of left-leaning PRI politicians who protested both the method of
selecting the next president and the economic model that the group in power had
adopted. The decade of the 1980s was marked by two extraordinary economic crises
that finally convinced President Miguel de la Madrid (1982–88)—with the support of
his economic team, led by the secretary of planning and budget, Carlos Salinas de
transformation of the dedazo in mexico 153
Gortari—to take initial steps toward liberalizing the Mexican economy. The leaders
of an antineoliberal faction within the PRI, the Democratic Current (Corriente
Democrática; CD) publicly protested de la Madrid’s right to unilaterally impose his
successor, in part because they were (rightly) convinced that the president would
choose the neoliberal Salinas to succeed him.
The regime leadership was quick to use old-fashioned methods of suffocating the
proto-rebellion, and offered many of those who had been present at the first meetings
selective incentives to leave the CD. Others left when they realized the course the
CD’s leaders were taking—a rupture with the regime. The regime’s other reaction to
the threat of a rupture was to organize a ‘‘beauty contest’’ among several possible
candidates, who gave speeches to several groups within the PRI and executive bu-
reaucracy. However, the pasarela (‘‘charade’’) fooled no one: President de la Madrid
would make his decision regardless of the opinions of his fellow PRI politicians. A
seemingly innocuous change in the electoral rules, however, now allowed multi-
party coalitions to back a single presidential candidate, and this allowed the now ex-
PRIistas to run against official PRI candidate Carlos Salinas in the 1988 election.∏
The 1987–88 presidential succession constituted the first serious breakdown in the
dedazo tradition since 1952. However, it was the electoral competition of the 1990s
that dealt the final death blow to presidential impositions.
Competition at the ballot box began to threaten the PRI’s hegemony in the
mid-1980s and grew sharply in the 1990s. This process changed the incentives of
politicians in the almost seventy-year-old governing coalition by making them less
likely to accept their defeats in nomination contests. When they had no exit option
because of the weakness of opposition parties and the government’s control over the
apparatus of elections, disgruntled PRI politicians had little choice but to accept the
will of the president (or governor, for local electoral posts such as mayors and state
deputies) and await a better political moment. With the growth of the electoral
popularity of party options other than the PRI, its politicians were strengthened: they
could again leave the coalition and run under another party’s banner and hope to
win the election. At the same time, as opposition parties battled to reduce the use of
fraud to change election results, there was a greater likelihood that electoral out-
comes would be respected. Electoral competition, which made party ruptures more
likely, helped end the dedazo.
In the 1993 succession process for the 1994 presidential election, then-president
Carlos Salinas Gortari played a traditional game in imposing his successor. After the
1987–88 rupture from the party ranks, it was of utmost importance that Salinas
control the succession, meaning that no disgruntled losing pre-candidate should
154 joy langston
leave the ranks of the PRI, form a new party, and run against the official candidate. In
late 1993, the PRI’s candidate was announced, and all losers accepted the presi-
dential decision (in one case, grudgingly). However, once the armed rebellion in
Chiapas began on January 1, 1994, the succession began to unravel. One of the losing
contenders, Manuel Camacho Solís, left the government to head up the negotia-
tions with the Zapatista rebels. Because he gave up his position more than six
months before the general election, Camacho was still legally able to participate as a
candidate. It was debated publicly whether Salinas second-guessed his initial deci-
sion by allowing Camacho to continue in the public eye without staying in his
government post. In March 1994, in a crime that still has not been fully solved, the
PRI’s presidential candidate was assassinated while campaigning. Salinas remained
strong enough to impose another close ally, the former secretary of planning and
budget Ernesto Zedillo, but the succession process was sullied: it was the first time
since 1928 that a presidential candidate had been murdered. The 2000 transfer of
power would end the dedazo forever. Unlike the presidents of the 1940s and 1950s,
there was little the PRI presidents could do in a fundamentally more evolved elec-
toral context to dampen the ambitions of its politicians and control their behavior.
The dedazo had been created under conditions of weak electoral institutions and
weak political parties in a largely rural nation. This situation had changed radically
by the 1990s.
In the 1990s, PRI politicians operating under newly competitive conditions began
to use the formal statutory rules to revoke the president’s control over the presidential
succession. There were two steps in the process to remove power from the hands of
the president. The first step was new statutory rules that stripped President Zedillo
(1994–2000) of his control over the universe of choice by forcing presidential candi-
dates to have prior electoral experience. This rule opened up the ranks of possible
PRI presidential candidates to many outside the cabinet, while disqualifying several
secretaries inside it. This rule not only remained on the books but was also respected.
The second step was a radical change in the procedure to choose the presidential
nominee; instead of a party assembly that simply ratified the president’s nominee,
President Zedillo was obligated in 1999 to devolve the nomination decision to all
registered voters in an open presidential primary.π
The Zedillo era would see fundamental changes in relations between the presi-
dency and the PRI. Ever-rising levels of electoral competition, a severe economic
crisis in 1994–95, and Zedillo’s desire to negotiate reforms with both opposition
parties weakened the traditional ties between president and party. These new cir-
cumstances not only allowed statutory rule changes that favored lower-level party
transformation of the dedazo in mexico 155
actors, but also made it too costly for President Zedillo to revoke the changes that
directly harmed his interests in the presidential succession.
The Seventeenth National Assembly of the PRI was called after the shocking
electoral losses at the gubernatorial level in 1995 (and the resulting fall of the leader
of the party) and the dramatic economic crisis of 1994–95. The party assembly
constituted a confrontation between the president, on the one hand, and the gover-
nors and members of the party bureaucracy on the other. The challenge to limit the
president’s choice of his successor and strengthen the autonomy the National Execu-
tive Committee (CEN) and party bureaucracy came in the form of the requisites for
becoming a PRI candidate for president and governor. These requisites were crucial,
because no president since President Gustavo Diaz Ordaz (1964–70) had held
elective office before winning the nomination for the presidency. Zedillo’s favored
candidate was typical for the 1970–95 period: an economist who had never held
office but had risen politically through various posts in the executive bureaucracy.∫
For the governors (and other PRI politicians) to become potential candidates, they
had to revoke the informal rule that disbarred them from the internal competition.
With the 2000 presidential elections on the horizon, these actors used the National
Assembly to rework the formal statutes to their political advantage.
The fight over limiting the president’s prerogatives came in the form of a carefully
worded change in the statute to increase the requisites that politicians would have to
meet to be a presidential (and gubernatorial) candidate. Originally, the reform
proposal that was to be presented in the National Assembly in September 1996 stated
that presidential hopefuls should have been either a party leader or an elected
official. This would have made it more difficult, but certainly not impossible, for the
‘‘technocrats’’ serving in the president’s cabinet, who tend not to have experience in
elected posts, to meet the standards.Ω However, after an uproar in the statutory
working session of the assembly, the final wording of the document to come out of
the National Assembly was changed to specify prior experience as a ‘‘militant, and
party leader and elected official.’’ This left only a few of the president’s cabinet
secretaries eligible and created many more eligible candidates outside the cabinet,
including several governors who obviously met the electoral requirements. Thus, the
ability of the president to handpick his successor was considerably diminished, as
most of his cabinet members were disqualified because they did not have elected
experience. This forced President Zedillo to prepare his allies’ way by sending them
156 joy langston
to the Senate in 1997, by bringing already eligible politicians into leading cabinet
positions (which he eventually did), or by calling for another National Assembly to
remove the offending requirements.
During this period, rising electoral competition at the ballot box at the municipal
and state levels gave the PRI governors far more political weight. First, they were
responsible for winning the federal deputy elections in their states under far more
competitive circumstances. Second, ambitious PRI politicians who wished to be-
come governors now had an exit option if they were passed over in the nomination
battle. And finally, the governors were now in a position to argue that they were the
best possible candidates to defeat popular opposition candidates in the presidential
elections.
The president and his party leadership traditionally had veto power over the PRI
governors’ choices of their local deputies and municipal president candidacies, and
the PRI governors also had to negotiate with the center over what percentage of
federal deputy nominations they could place. As we saw above, the rule that only
cabinet members could be considered presidential candidates meant that sitting
governors were necessarily excluded. To remedy this situation, the governors and
party bureaucrats used the formal candidate selection rules to weaken the president’s
informal right to impose his successor. In abstract terms, a PRI governor would
prefer to end five decades of exclusion from the PRI presidential succession game
and win the informal right to be considered a candidate. To achieve these goals, he
would have to either modify the formal rules of the presidential nomination process
or pressure the president to give up his informal prerogative to name his successor, a
very unlikely scenario. Actors such as governors could attempt to reduce the universe
of presidential possibilities to exclude the president’s closest technocratic allies or
change the nomination method so the president could no longer simply impose his
successor.
President Zedillo was thus unable to simply overturn the formal statutory rules—
an astonishing change in ‘‘politics as usual’’ in Mexico. Zedillo’s institutional posi-
tion, which obliged him to negotiate with the opposition in Congress, and the
economic difficulties during his presidential term were some fundamental causes of
his weakness relative to the PRI and the PRI governors. Most importantly, the ever-
rising levels of electoral competition forced him to concede a radical change in the
presidential nomination process and, thus, an end to his traditional succession pre-
rogatives. It was better for him to attempt to win the nomination struggle within his
transformation of the dedazo in mexico 157
own party even under the new, more restrictive requisites than to risk an internal
split, such as those seen in the 1940s and in 1987. The changes made in the reform
assembly of 1996 and Zedillo’s failure to overturn these rules demonstrate two funda-
mental points: first, the growing importance of both the governors and the party
bureaucrats within the PRI; and second, the use of formal party rules as instruments
by weaker members of the coalition to shore up their position vis-à-vis the informal
prerogatives of the president.
Because the president could no longer guarantee electoral victories for his party
members, the collective good of assured victory was no longer provided in exchange
for loyalty and obedience. Thus, if an individual politician refused to remain loyal
when he was passed over for the nomination, then he could leave and run for
another party option, and there was no guarantee that the PRI would win the general
election. This raised the costs of staying loyal to the president, while at the same time
lowering the gains from remaining disciplined to the president’s mandates. Loyalty
and obedience to presidential impositions made sense when there was no chance
the PRI would lose in the present or future; now there were no guarantees.
The creation of formal requisites for presidential nominees was not the only
shocking change in the reigning balance between the formal and informal rules. In
1999 the covert imposition of the candidate ended when the president chose to
organize a primary of all registered voters to nominate the PRI’s presidential candi-
date.∞≠ This process took the selection out of his hands and gave it to the general
electorate, the rank and file (who were more likely to vote in the primary) and other
powerful actors within the party who helped organize support for the four candidates
at the state level.∞∞ Looking carefully at the statutes for choosing presidential candi-
dates, one finds that in 1999 there were no clear indications that an open primary was
included in the nomination methods (which included selection by the National
Assembly, made up of sectors and other groups within the party, or through a vote of
the National Political Council). The statutes were not changed in 1999 to explicitly
include a primary, because Zedillo did not wish to call another assembly (the party
body charged with statute reform). In yet another example of statutory ambiguity, a
primary was held under the ‘‘other method in special circumstances that the leader-
ship [read president] decides’’ clause of the article covering nominations (Party of
the Institutional Revolution 1993, art. 159).
On November 7, 1999, with ten million voters reportedly participating, the PRI
overturned seventy years of top-down decision-making within the party. Electoral
competition made the threat of a rupture of the PRI by one of the losing candidates
real and credible. This forced Zedillo and his advisors to find a procedure that would
be accepted as fair by all those competing for the nomination. None of the presiden-
158 joy langston
tial hopefuls would have accepted a party nominating convention or a vote of the
members of the National Political Council, because of the ease of predetermining
the outcome via the selection of delegates.∞≤ The open primary, despite the dangers
it represented in terms of fairness, was considered the best method available.
Due to the changes in levels of competition and the power this gave to other PRI
actors, several informal practices surrounding the presidential succession changed
in 1999. The president could no longer unilaterally impose his candidate without
risking a regime split, and this gave PRI leaders the power to change formal rules in
an explicit attempt to disqualify some of his favorites. Because the sitting president
no longer had complete control over the succession process, and because of the
exclusionary requisites, many prominent members of the PRI who were not mem-
bers of the cabinet were considered (or at least considered themselves) serious con-
tenders for the nomination. It became rational to openly state one’s presidential
ambitions and to mobilize support within the coalition for a run at the presidential
nomination. Powerful members of the PRI, such as Roberto Madrazo (then governor
of Tabasco), Manuel Bartlett (then governor of Puebla), Francisco Labastida (then
secretary of gobernación and Zedillo’s favorite), and Humberto Roque Villanueva (a
former leader of the PRI and congressional majority leader), declared that they
would compete, and warned Zedillo to avoid a regime rupture through more open
and democratic selection mechanisms. As a result, Madrazo, the strongest runner-up
to the eventual winner—Zedillo’s favorite—accepted his defeat and stayed within the
confines of the coalition.
conclusion
within and outside the coalition, or to create an electoral vehicle to challenge the
PRI’s official candidate at the ballot box.
These developments seem to show that the informal rules were competing with
some of the most important constitutional provisions that allowed for democratic
elections and party provisions that allowed for rank-and-file participation in choosing
candidates. However, this is too simplistic a representation. First of all, PRI regime
leaders had such utter control over the nation that they could have changed the
Constitution if they had chosen to. Informal rules were not so much competing with
the formal institutions as underlying them. Second, many formal rules, both elec-
toral and statutory, were in fact written to support the more basic informal preroga-
tives. The 1946 electoral reforms, and all of those until the early 1960s, concentrated
the organization and management of elections in the hands of the regime’s executive
bureaucracy and made party registration increasingly difficult. All candidates had to
represent a registered party, and independent candidates were barred from compet-
ing. And once the elections had taken place, the sitting Congress determined the
validity of the processes, and so could always approve any fraudulent behavior on the
part of the PRI’s electoral machine. Far from being a residual category, informal
institutions were ultimately the basis for political outcomes.
These informal rules were transformed half a century later, in large part because
of an external shock in the form of heightened electoral competition. The series of
electoral victories that began in the 1980s changed actors’ incentives and resource
base, which eventually led many of them to use the formal party rules to improve
their chances of winning candidacies or of curtailing the president’s ability to uni-
laterally impose his successor. The threat of rupture became far more credible in a
context of heightened competition—a paradoxical outcome, because opposition
parties had fought to rewrite offending electoral rules to ensure fairer electoral
processes, not to empower PRI party politicians. Once the electoral balance of power
had shifted, party actors used seemingly insignificant formal rules to quickly trans-
form informal practices that had endured for decades.
chapter eight
One of the most striking elements of Chilean politics in the years since the return
to democratic government in 1990 is the stability of the governing Concertación
coalition. Originally founded in the 1980s as a broad amalgamation of parties and
civic groups in opposition to the perpetuation of General Augusto Pinochet’s military
regime, the Concertación evolved into an electoral and governing alliance composed
of four major parties—the Christian Democrats, the Socialists, the Party for Democ-
racy, and the Radical Social Democrats—along with a handful of independents, that
has endured through three presidential elections, four legislative cycles, and four
cycles of municipal contests. This stability stands in striking contrast to the coalitional
fluidity that characterized Chile’s long pre-Pinochet democratic experience (A. Val-
enzuela 1994, 120–25). Chile’s current governing alliance bears resemblance to gov-
erning coalitions in parliamentary systems, which are frequently characterized as
more durable than legislative coalitions in multiparty presidential systems, the latter
frequently characterized as improvisational, circumstantial, and devoid of recogniz-
able policy content (Mainwaring 1993; Linz 1994). As a result, Chile’s coalition ought
to be immune from the ills frequently attributed to loose presidential coalitions, in
particular that they undermine accountability by making it difficult for voters to
identify which parties share in responsibility for government performance.
The formal institutions handed down by the Pinochet regime during Chile’s
transition to democracy have, by some accounts, contributed to coalition stability
(Rabkin 1996; Carey 2002; Agüero 2003). In particular, the unique rule by which all
elected legislative seats are contested in two-member (M=2) districts encourages
formal and informal institutions in chile 161
coalitions in that, in any given district, only candidates from the top two lists can win
representation. At the same time, however, the M=2 rule implies that to secure
legislative majorities, coalitions must nominate their strongest candidates in the most
precarious electoral list positions. This conundrum can generate substantial tension
within coalitions, particularly one as broad as the Concertación, in which nomina-
tions must be divided among many actors (Siavelis 2002b).
In this chapter, we argue that the Concertación has responded to the challenge
implied by the M=2 formal institution by creating and sustaining an informal
institution of insuring strong candidates who incur risk on behalf of the coalition
against the vagaries of the electoral marketplace. Specifically, there is an informal
system of rewarding good also-rans with appointed government posts. The mutual
expectations generated by this arrangement, on the part of politicians and party
leaders, have helped to recruit strong candidates and sustain the Concertación. At the
same time, recent changes in the Chilean electoral landscape threaten Concertación
control over the pool of resources that indemnify its best candidates against electoral
risk, and these changes may therefore threaten the viability of the coalition itself.
The Concertación’s demise has been predicted often over the decade and a half
during which it has governed. Even the coalition’s leaders have issued premature
announcements of its death (Latin America Adviser 2002; El Mercurio 2003a). Some
observers of Chilean politics, therefore, might regard the dissolution of the Concer-
tación as both a foregone conclusion and an overdetermined event—that is, as the
product of so many factors, all working toward the same end, that it is impossible to
assign responsibility among them. We share the opinion that the Concertación may
not endure much beyond Chile’s 2005 election, but not for the reasons commonly
cited. Our argument is based, instead, on the structure of Chilean political careers,
which in turn is connected inextricably with Chile’s unusual electoral rules.
Those foreseeing the Concertación’s demise point out that the Chilean economy
slowed in the early years of the twenty-first century, after the coalition’s first years in
government during the 1990s. Moreover, the coalition itself was initially galvanized
around opposition to the Pinochet regime of the 1970s and 1980s, so as time passes,
the compelling force of that initial motivation might naturally weaken. In addition,
the Concertación, and even its component parties—most notably the Christian
Democrats—are internally divided over social issues, such as the legal status of
divorce and access to birth control. Finally, by the 2005 elections, the Concertación
will have held the presidency and a majority in the Chamber of Deputies (the
Figure 8.1. Sample Ballot from 1997 Election for Chamber of Deputies
formal and informal institutions in chile 163
legislative house elected by popular vote) for sixteen years. By the standards of
multiparty coalitions anywhere, much less among Latin America’s presidential sys-
tems, the Concertación is geriatric, bearing the scars of miscellaneous corruption
charges against members—including the stripping of parliamentary rights from five
of the coalition’s deputies in 2002—and the general disillusionment that goes with
holding the reins of power for so long. One might conclude, then, that the coalition
is simply ready to expire.
Any of these forces, or some combination of them, could indeed undermine the
Concertación, but we do not regard these factors as necessarily devastating to the
coalition’s survival, for a number of reasons. In the first place, the Chilean economy
came through the lean years at the turn of the century in far better shape than that of
any of its Southern Cone neighbors, and employment and growth figures rebounded
in 2003-4, providing the coalition with a plausible claim to good economic steward-
ship. Next, while internal differences on social issues plague the coalition, this has
been the case since it began to govern, and similar divisions bedevil its main oppo-
nents on the right (here designated the Right). Finally, who is to say how old is too
old for a coalition? Based on Chile’s historical experience, the Concertación had
already far exceeded the life expectancy of any multiparty coalition as early as the
1997 legislative elections, yet it prevailed in that instance, and in the 1999–2000
presidential election, and yet again in the 2001 legislative elections (Carey 2002).
To sum up, many of the conditions widely regarded as threatening to the Concer-
tación’s survival are not new. They have been present during much of its lifetime, and
so represent constants in the electoral environment. Moreover, as many observers
have noted, Chile’s electoral system presents substantial risk for parties that abandon
coalitions and run on their own. Indeed, the explanation we advance here regarding
the Concertación’s potential demise rests on the unique nature of Chile’s electoral
law and how this interacts with recent changes in the electoral environment.
table 8.1
Summaries of District-Level Coalition Competition in Chilean Congressional Elections,
1989–97
Concertación
Concertación Concertación wins one seat,
Right doubles wins neither plus non- Concertación
doubles Right seat Right winner runners-up
Chamber
1989a 0 11 2 1 40
1993 1 11 1 0 48
1997 0 9 0 4 47
Senate
1989b 0 3 0 0 16
1993 0 0 0 0 9
1997 0 1 0 0 9
a
In six districts in 1989, the Concertación ran only a single candidate for the Chamber of Deputies, leaving no
runner-up. Thus the row does not add up to sixty, the total number of districts.
b
In 1989, all thirty-eight elected Senate seats (nineteen districts) were up for grabs. In subsequent elections,
alternately nine or ten districts are contested every four years.
seats to lists, then seats are awarded to individual candidates in the order of their rank
within their list. Seats are allocated by the D’hondt method, such that the first-place
list in a district can win both seats only if it more than doubles the vote total of the
second-place list; otherwise, each of the top two lists wins one seat.≤
Though the electoral rule is procedurally straightforward, its combination with
the contemporary party system generates substantial strategic complexity for candi-
dates, parties, and coalition leaders. Electoral politics in postauthoritarian Chile has
been dominated by two major coalitions: the Concertación on the left and the
Alianza por Chile on the right—each of which in turn is composed of two subpacts.
For most of this period, the Concertación has been composed of four significant
parties (PDC, PRSD, PPD, PS) and the Alianza por Chile has consistently included
two major players (RN, UDI), along with some minor regional parties.≥ Indepen-
dents associated with particular ideological sectors play an important role and have
also had to negotiate themselves onto major lists.
The constituent parties of each coalition must negotiate sixty two-seat electoral
slates for the Chamber of Deputies and either nine or ten two-seat slates for the
Senate, depending on the cycle of turnover in the upper house. The pattern in past
elections has been for each of the coalitions to allocate one seat to each of its two
subpacts in each district, and then to tinker with list formation around the edges to
attract small parties that can lend a helping hand in bolstering national coalition
vote totals for presidential elections.
Two stark realities shape negotiations. First, most lists can expect one defeat in
each district, because the threshold for two-seat victories is so high. Table 8.1 summa-
formal and informal institutions in chile 165
rizes the number of two-seat victories in recent elections for each major coalition.
Second, and more important for our purposes, the pairing of candidates is crucial.
The key to victory is for subpacts and parties to place their candidates on the same list
either with an extremely strong candidate, who will help them more than double the
vote total of the second-place list, or a relatively weak candidate, who will not outpoll
their own candidate while the list pulls enough votes to win one seat. This deli-
cate balance creates a tension between the preference of candidates and coalitions.
Coalitions would like sixty pairings that maximize votes, whereas candidates may
often prefer a weak list partner.∂ Those who are outpolled by their list partner in these
two-seat pairings we refer to as ‘‘runners-up.’’ We explore their fate below.
precisely the strongest candidates—the ones whose personal qualities and popularity
among voters make them best able to win elections—that coalitions need to put in
the riskiest positions in order to pursue the collective goal of doubling and win
legislative majorities.
A strong candidate has every reason to prefer to be paired on a list with a relatively
weak partner in a district where her coalition is willing to settle for the one-one split
between first- and second-place lists most common in Chilean elections. To be the
strong candidate on such a list is, effectively, a guarantee of personal victory. To be
nominated in a district where one’s coalition aspires to ‘‘double,’’ on the other hand,
means facing not only competition from other lists but also competition from one’s
list partner. The very imperative that leads coalitions to run strong pairs of candidates
in districts where they seek to double threatens the electoral security of their best
politicians.
How has the Concertación resolved this conundrum? We suggest that to induce
strong candidates to embrace the risk inherent in a doubling campaign, the coalition
has offered insurance. In the language of political economy, for over a decade the
Concertación has reaped substantial profits from its ability to act collectively, and
the currency of those profits is control of the government—both of the leadership
positions in Congress, due to the Concertación’s success in doubling its opponents
in key districts, and of the executive, due to the coalition’s ability to nominate
presidential candidates and support them in a unified manner. For strong legislative
candidates who take risks on the coalition’s behalf by attempting to double, insur-
ance takes the form of a promise of attractive appointed positions in the government
if one should fall short in the electoral competition.
The participants in this informal institution are potential candidates for nomina-
tion to legislative lists, and the coalition leaders who recruit candidates and negotiate
the allocation of nominations among parties participating in the coalition. Coalition
leaders, in this case, are the leaders of the Concertación’s component parties—who
may not necessarily hold elected office—in conjunction with the parties’ top legisla-
tive leaders, aides of the president of the Republic, representatives of the Ministry of
the General Secretary of the Presidency, and members of the partido transversal.
The partido transversal is an informal, though well-recognized, cadre of supra-party
elites whose loyalty lies as much with the Concertación coalition as with their
individual parties. The partido has been crucial in generating the many agreements
formal and informal institutions in chile 167
that hold the coalition together (Fuentes 1999; Walker 2003; Siavelis, this volume).
These sets of actors strike deals in informal and multilevel negotiations. In the most
basic terms, the deal with respect to legislative candidates is: ‘‘You put yourself at risk
in this election for the good of the coalition. If we win both seats, terrific. If we fail to
double and you lose out despite having done your best on behalf of the coalition, you
will be compensated with a ministry or some other attractive executive appoint-
ment.’’ The deal, of course, is not a written contract but an informal agreement on
which the mutual expectations of candidates and coalition leaders are based. In
conversations with legislators from both inside and outside the Concertación, we
have confirmed that such an expectation exists among politicians.∏
Our goal here is to identify the conditions under which runner-up candidates on
Concertación lists that failed to double are appointed to top government posts dur-
ing the four-year period in which, had they won, they would have served in Con-
gress. Our hypothesis is that ‘‘good losers’’—those that competed valiantly in defeat—
are the most likely to receive appointments, and in a manner consistent with the
informal institution of insurance for candidates that incur risk for the good of the
coalition. The dependent variable, then, is whether the runner-up received an ap-
pointment as a minister, vice-minister, or ambassador during that time. We posit
that the factors that should affect the probability of appointment are: for which
chamber the runner-up ran, how narrowly his or her list failed to double the next
closest list, and how evenly the personal votes for candidates were split within the
Concertación list.
Senate races are more important than Chamber of Deputies races because,
although the chambers have analogous powers, there are fewer seats in the upper
chamber, rendering each seat more precious. The Senate is also perceived as the
launching pad for presidential candidacies. Because the stakes are higher, electoral
competition for the Senate is fiercer, so strong performance is more impressive. Also,
the candidates recruited for Senate races are more prominent. This has two implica-
tions. First, their experience and qualifications mean these candidates are often
‘‘ministerial caliber.’’ Second, high-credential Senate candidates have attractive out-
side options, and to induce such individuals to undertake a risky campaign, the
insurance for runners-up must be more generous than that for Chamber candidates.
For all these reasons, we expect senatorial good losers to be more likely to receive
appointments than Chamber good losers.
The closer the Concertación came to doubling the Right,π the more impressive
was the collective performance of the list in its attempt to double. Our expectation is
that runners-up from such lists should be increasingly likely to be rewarded for their
contribution to the collective effort.
168 john m. carey and peter siavelis
How evenly matched the two candidates on the Concertación list are in terms of
personal votes is a more subtle issue. If a strong list performance is largely the
product of the vote-attracting capacity of one candidate, then the runner-up may be
regarded as not instrumental to the list’s overall performance, and therefore not as
deserving of postelection reward as when the list’s two candidates ran neck and neck.
By this logic, runners-up from ‘‘lopsided’’ lists should be less likely to receive appoint-
ments. Yet it is not necessarily in the coalition’s collective interest to foster intra-list
competition. That is, a coalition benefits from having strong candidates who attract
votes, in a generic sense, to the list, but it does not profit from a competition by
which list-mates seek to take votes from each other. Indeed, too much internecine
competition could prove damaging to a list’s overall status and depress its overall
votes, much as parties in the United States worry about the effects of a tough primary
campaign on the party’s eventual candidate. In order to mitigate incentives for
‘‘cannibalistic’’ intra-list competition, then, the coalition may want to emphasize
collective list performance in providing insurance to runners-up.
In short, we expect a higher probability of the Concertación runners-up getting
an appointment (1) in Senate, rather than Chamber, races; (2) when the Concerta-
ción list beats the Right more severely (approaching a doubling); and (3) (with less
confidence here) when the two Concertación candidates contribute more equally to
the list’s overall vote total. We would interpret such results as evidence that the
‘‘insurance policy’’ of appointments by the executive was wielded systematically as a
way of attracting strong candidates to fight tough races and rewarding those that did
so valiantly in a losing cause.
At this point, it is worthwhile to emphasize what we regard as the relevance, and
the limitations, of our statistical analysis. Our model measures whether, among
runners-up, stronger candidates are more likely than weaker candidates to receive
appointments. Our measure of candidate strength is based entirely on electoral
outcomes, which reflect combinations of two types of characteristics: (1) inherent
quality (charisma, qualifications, personal prominence) and (2) effort on behalf of
the campaign. The logic by which these two types of qualities contribute to the
probability of appointment differs. We expect that a potential candidate’s inherent
strength raises the opportunity costs of accepting a risky electoral list position, be-
cause the same qualities that make candidates inherently strong also increase the
odds that they have career options more attractive than placing themselves at elec-
toral risk. For these types of candidates, a promise of insurance is critical to entice
them to accept a spot on a high-risk (attempting-to-double) list. With respect to a
candidate’s effort, the logic is more straightforward. All else being equal, the presi-
dent should be more inclined to reward those who worked hard on behalf of the
formal and informal institutions in chile 169
coalition and thus fell just short of doubling than those who campaigned less effec-
tively and so fell further short of the mark. Although there are two distinct logics at
work here, both generate the same expectations with respect to the parameter esti-
mates in our model.
Ideally, we could know inherent candidate quality separate from candidate effort,
and thus separate out the independent effects of each characteristic. If we could do
this, then a relevant comparison would be between Concertación winners from lists
that elect only one candidate (i.e., those that do not double) and Concertación
runners-up in districts that the coalition targeted to attempt to double. If we are cor-
rect, then good losers from attempting-to-double districts should be of even higher
inherent quality than winners from the full set of districts in which only one candi-
date from the Concertación was elected. Unfortunately, we do not have a way of
measuring inherent candidate quality distinct from electoral outcomes. In the end,
then, we acknowledge that our analysis does not allow us to disentangle the ‘‘inher-
ent quality/opportunity costs’’ story about insurance from the ‘‘reward for effort’’
story. In fact, of course, these two stories are mutually consistent, and we think there
is truth to both of them.
evidence
1. We identified the potential Concertación good losers, who are the coalition’s
runner-up candidates in districts except where the following conditions apply:
≤ the Concertación doubled the Right, thus electing both candidates on its
list to Congress (i.e., no losers);
≤ the Concertación list included only one candidate, who won a seat;
≤ the Concertación won neither seat, thus suffering electoral humiliation in
the district; or
≤ the Concertación won the first seat, but the second was won by a candidate
from a list other than the Right, such that the Concertación failed to
double even against a divided opposition—the conditions that ought to be
most propitious for doubling. (Table 8.1 shows the frequency of each of
170 john m. carey and peter siavelis
these scenarios, as well as the remaining number of races from each elec-
tion (in the column ‘‘Concertación runners-up’’)—races that produced
potential good losers and so comprise our data set.)
2. We collected the names of those appointed to a set of political posts during the
period 1990–2001—the legislative periods following the three elections we
analyze.
3. We matched the names of our runners-up with those who received a plum
appointment during the legislative period following their electoral defeat.
Table 8.2 categorizes the positions for which we searched and our sources. The
Chilean president has wide powers of appointment, naming approximately thirty-
five hundred government officials (El Mercurio 2003b). We focus only on high-
profile posts. Lower-level positions are less attractive to ambitious politicians and are
less likely to be used as rewards. We certainly do not include all important appoint-
ments. There are other posts with which presidents might reasonably reward politi-
cians who served valiantly in their coalition’s electoral battles for which we were
unable to obtain systematic data (e.g., boards of state-run corporations, regional
governorships). Beyond these, there are still other positions (e.g., judicial appoint-
ments, party offices, future nominations to congressional or municipal council slates
of candidates) that are not direct presidential appointments but over which presi-
dents could exercise considerable influence on behalf of their allies. In short, our
search for appointees almost certainly includes some Type II error, but not Type I
error. That is, we almost certainly miss identifying some good losers who received
appointments from the executive, or at the behest of the executive, but we are certain
that all those we identify as having received appointments are correctly classified.
Our second strategy for collecting data on appointments complements the first by
minimizing Type II error (failing to detect appointments that were made) through
intensive efforts to map the career trajectories of runners-up following their electoral
losses. Because we could not conduct such exhaustive efforts for all candidates in the
study, we limited this strategy to those Concertación good losers who were most and
least likely candidates for appointment, according to our hypotheses. We identified
‘‘most likelies’’ as those runners-up on lists in which the Concertación-to-Right vote
ratio was greater than 1.8—that is, the list won more than 90 percent of the necessary
share of the vote to be able to ‘‘double’’ the Right, falling just short. There were
twenty such good losers: fourteen candidates from the Chamber of Deputies and six
from the Senate. We identified ‘‘least likelies’’ (‘‘bad losers’’?) as runners-up on lists
that received fewer votes than the coalition of the Right (Concertación-to-Right ratio
between 0.5 and 1.0). This yielded nineteen Chamber runners-up and only one from
formal and informal institutions in chile 171
table 8.2
Available Positions Included in Appointment Variable, Data for 1990–2001
Position Number of available posts Sources of data
the Senate. So as not to exclude the Senate among ‘‘least likelies,’’ we added the next
four least successful Concertación Senate runners-up (with ratios up to 1.13).Ω
For our set of ‘‘most and least likelies,’’ we supplemented the data included in
our systematic search for appointments with standard internet searches; ‘‘manual’’
searches of Chilean government, party, newspaper, and nongovernmental organiza-
tion websites and archives; and email correspondence with Chilean academics and
government officials, inquiring about the postelectoral careers of specific runners-
up. This approach yielded information on some executive-level appointments to
positions for which we do not have comprehensive data, as well as information on
some appointments to other posts, but for which presidential support may have been
instrumental.∞≠
As a first cut at the question of whether Concertación good losers were systemati-
cally rewarded with appointments, consider the results of the exhaustive search
approach for appointments to ‘‘most and least likelies’’ (summarized in table 8.3). We
identify desirable postelectoral appointments for 70 percent of our ‘‘most likelies,’’ but
for only 25 percent of our ‘‘least likelies.’’ These include presidential appointments
during the immediate post-good-loser period for three of our six ‘‘most likely’’ Senate
good losers, plus prominent public posts not directly appointed by the president for
two others, and nomination to a (successful) senatorial bid for the sixth. Thus, all six
‘‘most likely’’ Senate runners-up prospered. Among the five ‘‘least likely’’ Senate
runners-up, we identified a presidential appointment during the subsequent period
for one, during the second subsequent period for another, and nothing for the other
three. There are more Chamber runners-up, but the pattern is the same: we identify
subsequent posts for most of the ‘‘most likelies,’’ but nothing for the ‘‘least likelies.’’
The manner in which we selected cases for exhaustive career-trajectory searches,
of course, should be expected to generate a bias toward such stark results. What we
can conclude at this point is that those runners-up we have posited to have the most
promising post-electoral-loss prospects do, in fact, have good odds of receiving plum
appointments (perhaps even better odds than the table shows, given that we may still
172 john m. carey and peter siavelis
table 8.3
Postelectoral Career Trajectories for Runners-up Deemed Most and Least Likely to
Receive Appointments, According to Their Lists’ Concertación-to-Right Vote Ratios
Post-runner-up appointment Most likelies Least likelies
have missed some appointments), whereas for those whose prospects we posited as
least promising, the odds were a lot worse.
In other work (Carey and Siavelis 2005), we test a logistic regression model of the
conditions under which second-place candidates receive appointments, drawing on
all Concertación lists during the first three electoral cycles since Chile’s transition to
democracy in 1990. Here, we briefly describe the analysis and its results.
formal and informal institutions in chile 173
Interviewer: Is there an agreement, for example, where if you lose this election, then
you get a post?
UDI Deputy: Certainly, yes. In practice, everyone—above all, in the government,
which has more resources to give positions than the opposition . . . in the opposi-
tion, it’s less—receives a post within the party or some compensation. But it’s not
very centralized—for example, some businessman has a job . . . In the government,
there are a lot more resources to provide positions, either in the government itself,
in foreign relations, embassies, which are very attractive.
Interviewer: Are there agreements ahead of time?
174 john m. carey and peter siavelis
UDI Deputy: I don’t know if it’s so explicit, but certainly I believe that there is an un-
derstanding. The Concertación has been pretty efficient—probably too efficient—
in keeping its leading politicians inside the system, you know? They haven’t lost
important leaders. And they’ve been able to keep them in the government, in Con-
gress, in party positions, in public corporations or foreign relations posts.∞∞
Another UDI deputy elaborated further, suggesting that insurance in the form of
appointments to government posts has been more important for the governing coali-
tion than it would be for the Alianza, because candidates of the center-left parties
have fewer personal resources and so rely more heavily on public sector employ-
ment.∞≤ Finally, we note that the public reaction to an early version of the argument
in this chapter, published in Chile in 2003, suggests widespread acknowledgment
of the insurance system by politicians and political observers (El Mercurio 2003c;
El Sur 2003).
The Chilean electoral system generates a tension between the incentives for
coalitions and those for individual candidates. To align politicians’ incentives with
those of the broader coalition, the Concertación has provided insurance for its good
losers. The bottom line from our analysis is that those who were recruited for the
most prominent (i.e., Senate) races and those whose efforts most nearly bore fruit
(i.e., whose lists nearly doubled those of the Right, thus contributing to Concerta-
ción legislative majorities) were the primary beneficiaries of the insurance-through-
appointments. This evidence suggests, then, that the Concertación has used politi-
cal appointments to reconcile the interests of individual politicians with those of the
coalition as a whole—to reward politicians who accepted personal risk on behalf of
the coalition and who contributed toward the collective performance of their lists—
and thus to overcome the divergence of incentives generated by Chile’s unique
electoral system.
This insurance system has worked well for the Concertación, and for its candi-
dates, in each of the four elections since the transition to democracy in Chile. Why
might it fail now? We suggest that the insurance claim fund is no longer as secure,
because it depends on control of the executive branch. The danger, from the per-
spective of would-be doubling candidates, is that the Concertación may not control
the executive branch after the 2005 election. The possibility began to crystallize in
1999–2000, in the form of Joaquín Lavín’s stronger-than-anticipated challenge to
Ricardo Lagos for the presidency. After that initial campaign, Lavín positioned
formal and informal institutions in chile 175
In 2005, the situation will be different. Viable presidential challengers from the
Right (unlike in 1989 or 1993) mean that the coalition’s ability to pay insurance to
176 john m. carey and peter siavelis
‘‘doubling losers’’ will be less than certain. In 2005, moreover, legislative and presi-
dential elections will be concurrent for the first time since 1993. The 1997 and 2001
legislative elections were held in the middle of presidential terms. The existence of a
concurrent presidential election in 2005 adds an additional dimension of uncer-
tainty for coalition leaders and for prospective legislative candidates, because bar-
gaining over the composition of lists for the legislative election will take place before
the Concertación’s presidential candidate is determined by a primary election. Fac-
ing uncertainty about compensation in the event of an unhappy outcome, the
question is whether strong candidates, and their associated parties, will remain com-
mitted to maintaining the coalition.
In the terms set out by Helmke and Levitsky in their introduction to this volume,
the insurance system for good losers in Chile’s Concertación is a complementary
informal institution in that it serves to compensate for a dilemma generated by a
formal institution that might otherwise undermine political cooperation. The di-
lemma is the incentive in Chile’s two-member district elections for coalitions to
place their strongest candidates in their most electorally vulnerable races. Insurance,
as an informal institution, compensates candidates for the risk they take on, and
thereby helps maintain the coalition. Helmke and Levitsky argue persuasively that
among informal institutions, complementary ones ought to be particularly respon-
sive to changes in formal institutions, on the grounds that if the complementary
institution took shape to compensate for some characteristic of the formal institu-
tional environment, we might reasonably expect it to wither, or adapt, if the formal
institution it complements is altered. We find Helmke and Levitsky’s logic com-
pelling here, but in the case of insurance for good losers, we suggest a potential
wrinkle—that change in the informal institution may encourage reform of its formal
complement.
We have argued that the two-member district electoral system and insurance for
good losers are mutually reinforcing in Chile, but also that the latter is sustained by
the Concertación’s electoral dominance and may be threatened if such dominance
is in jeopardy. It follows that if the system of insurance is undermined, the strategic
dilemma inherent in two-member district elections reasserts itself. Of course, the
dilemma is always present for those outside the dominant coalition, but in a political
environment where all actors face uncertainty about who will control future ap-
pointments, such that even the majority coalition can no longer be indemnified,
lawmakers may find the two-member district electoral rule itself increasingly dis-
formal and informal institutions in chile 177
tasteful.∞≥ A more even electoral competition between Chile’s two major coalitions,
then, may increase support among legislators for a return to large-magnitude pro-
portional electoral rules more closely resembling those of Chile’s mid-twentieth-
century democracy. In this case, we would suggest that not only might changes in
formal institutions foster changes in complementary informal institutions, but the
reverse may be true as well.
Our findings also have cross-national theoretical applications, especially as po-
litical systems around the world increasingly adopt complex electoral systems in
an attempt to balance representation and stability. We think it plausible that in
other systems, compensation mechanisms could exist for losing candidates who bore
risk on behalf of their coalitions. Insurance systems may be less critical to coali-
tion maintenance elsewhere, because, under all other electoral systems with which
we are familiar, electoral economies of scale are more straightforward than under
Chile’s M=2 system. That is, the efficiency with which votes are converted into seats
under higher-magnitude systems means that the need to put strong candidates in the
most marginal list positions is mitigated. Still, findings from Chile suggest that
compensation systems of the sort resulting from the two-member system are likely to
develop in other strategically complex electoral systems. In terms of future theoriz-
ing about both electoral systems and informal institutions, it is crucial to analyze the
extent to which strategic complexity in electoral systems can lead to the creation of
informal institutions to compensate for uncertainty.
We want to emphasize two last points in closing. First, we do not suggest that the
good-loser system is the only, or even the most important, source of unity for the
Concertación. Our point is that the system provides an important coalitional adhe-
sive overlooked by analysts of Chilean politics, and the demise of this reward system,
along with other tensions emerging in the coalition, can significantly hasten the
dissolution of the Concertación. Second, we are not advocating the demise of the
Concertación, nor are we prepared to applaud this result, if it should come to pass.
In our estimation, the Concertación governments and legislative majorities have,
since 1990, provided Chile with some of the most enlightened and temperate politi-
cal leadership in the country’s history. Rather, our guarded opinion about its pros-
pects is based entirely on what we see as a fundamental change in the Chilean
electoral landscape since 2000, and what we understand to be the effects of this
change on the strategic nature of coalition formation and maintenance.
chapter nine
well to Latin America. Although many Latin American parties are formally struc-
tured along the lines of their European counterparts, there is often a vast gap be-
tween how those parties are organized on paper and how they function in practice.
In many cases, decision-making power lies not in formal leadership bodies but in
individual leaders or office-holding party bosses; career paths are determined not by
bureaucracies but by personal and patronage networks; local organizations do not
take the form of branch offices but rather are informal networks that operate out of
state agencies or activists’ homes; and legal party finance is dwarfed by informal (and
usually illicit) forms of finance, such as patronage, kickbacks, illegal donations, and
embezzled state resources.
These differences have important implications for how we study parties. If the
bulk of a party’s organization lies at the margins of its formal bureaucracy, or if party
work is carried out by members of a soccer fan club rather than official members,
then measurements of party organization that examine only official branches or
membership data will seriously understate the size or density of the organization.
Similarly, if decisions, resources, and career paths pass through informal networks
rather than a party bureaucracy, then analyses that focus only on formal structures
will produce a flawed understanding of how that party functions. Indeed, there is
growing evidence that informally organized parties behave differently from their
more formal counterparts in areas such as electoral and legislative behavior, candi-
date selection, and adaptation to environmental change.≥
Drawing on ethnographic research on party organizations in Argentina and Ec-
uador, particularly the Argentine (Peronist) Jusicialista Party (PJ) and the Ecua-
dorian Roldosista Party (PRE), this chapter takes an initial step toward conceptualiz-
ing and measuring informal party organization. The chapter develops indicators of
formality and informality in nine areas of party life: internal rules and procedures,
decision-making bodies, central bureaucracies, grassroots infrastructure, organiza-
tional boundaries, career paths, party membership, ancillary organizations, and fi-
nance. These areas cover both institutions as defined in Helmke and Levitsky’s
introduction to this volume (e.g., rules regarding membership, leadership selection,
and finance) and entities that are generally treated as organizations (e.g., local
branches and ancillary organizations).
Most of the informal party structures discussed in this chapter fall into one of two
types. Some (charismatic decision-making norms, illicit channels of finance) are
competing institutions, in that they run directly counter to party statutes and/or state
law; others, such as informal branches that perform such party work as activist
recruitment and voter mobilization, are best characterized as substitutive. Informal
180 flavia freidenberg and steven levitsky
structures thus thrive where formal party organizations are weak—where the guide-
lines laid out in party statutes are not fully implemented or enforced.
Formal party organizations are those that are officially sanctioned. They are
created through established party channels, usually according to guidelines estab-
lished by party statutes, and are recognized by official party (and often state) authori-
ties. This includes official headquarters, bureaucracies, and local branches or cells.
Informal party organizations, by contrast, are those that carry out partisan functions
without official sponsorship. Examples include personal, clientelistic, and patronage
networks, as well as civic and social organizations that are not affiliated to parties but
nevertheless engage in partisan work. Informal structures are not found in party
statutes and may not be recognized by party or state authorities. However, if we
define political party as ‘‘any political group that presents at elections, and is capable
of placing through elections, candidates for higher office’’ (Sartori 1976, 64), then
informal structures must be considered elements of party organization.
Informal party organizations should be distinguished from weakly institutional-
ized organizations. Party institutionalization, which may be defined as a ‘‘process by
which a practice or organization becomes well established and widely known, if not
universally accepted’’ (Mainwaring and Scully 1995, 4), is often conflated with for-
mal institutionalization. For example, Panebianco equates institutionalization with
an increasing ‘‘correspondence between a party’s statutory norms’’ and its ‘‘actual
power structure’’ (1988, 58–60). According to this conceptualization, all organiza-
tional patterns that depart from the ‘‘statutory norms’’ are noninstitutionalized. Yet,
as this volume makes clear, informal organizational activity may also be institu-
tionalized. For example, clientelistic or machine parties are often based on estab-
lished patron-client networks that operate with widely known and accepted norms
and procedures. Indeed, many of the most highly institutionalized parties in Latin
American history, including the Mexican PRI (Party of the Institutional Revolution),
the Colorados in Paraguay, and traditionally dominant parties in Colombia and
Honduras, are predominantly informal.
Distinguishing between formalization and institutionalization allows us to cap-
ture three distinct party types: (1) formally institutionalized parties, in which rules
and organizational structures are institutionalized in line with party statutes; (2) in-
formally institutionalized parties, in which formal structures are weak but informal
structures are well established; and (3) weakly institutionalized parties, in which
informal institutions and party organization 181
neither formal nor informal structures are well-established. Although both infor-
mally institutionalized parties (e.g., patronage-based machines) and weakly institu-
tionalized parties (e.g., personalistic or charismatic parties) are informal, in that their
formal structures are largely decoupled from the real organization, the two types are
very different and should not be conflated.
All parties contain elements of both formal and informal organization. Personal
networks and informal factions exist even in the most bureaucratic of parties. The
difference lies in the relative weight of formal and informal structures. In formally
organized parties, bureaucratic structures predominate. Decision-making, resource
allocation, and career advancement take place largely through official channels.
Informal networks operate within the parameters of the bureaucracy and do not
seriously compromise its effectiveness. By contrast, in informally organized parties,
decision-making, resources, and careers pass through informal structures at the mar-
gins of the bureaucracy—often rendering it ineffective.
As an initial step toward more rigorous measurement and comparison, this sec-
tion develops a set of indicators of formality and informality in nine areas of party life:
(1) internal rules and procedures, (2) locus of authority/decision-making bodies,
(3) central bureaucracies, (4) local organization, (5) organizational boundaries,
(6) party hierarchy/career paths, (7) membership, (8) ancillary organizations, and
(9) finance.
Where party organization is informal, behavior conforms less closely to the rules
laid out in the statutes. Formal rules are routinely violated or ignored (often by party
leaders themselves), and, due to a lack of effective enforcement, rule infractions are
rarely punished. As a result, the formal rules—and rule-making processes—are taken
less seriously. Some informal party organizations are characterized by the absence of
any (formal or informal) set of stable or binding rules. Thus, in charismatic parties,
‘‘no accepted procedures exist, and improvisation is the only real organizational
‘rule’ ’’ (Panebianco 1988, 146). In other cases, such as patronage-based or machine
parties, behavior may be guided by informal rules that, while diverging from party
statutes, are nevertheless widely known, accepted, and complied with. For example,
in the Mexican PRI before the 1990s, a variety of informal rules and procedures—
including the dedazo system of candidate selection—were highly institutionalized
(Langston, this volume).
Where party structures are formal, actors seeking to advance their careers in the
organization must follow certain specified—and generally written down—rules and
procedures. Leadership posts are filled through formal, publicly known procedures,
such as congresses or primaries. Often, procedures exist to filter out newcomers. For
example, the statutes of many mass bureaucratic parties stipulate that members must
belong to the party for a specific number of years before holding leadership posts, or
that leaders ascend through the hierarchy by means of sequential election to local,
midlevel, and then national leadership bodies. In many cases, party members cannot
be elected to a national leadership position without first being elected to a lower-
level body.
Where party organization is informal, career paths do not conform to formal
rules. Bureaucratic procedures either do not exist or are not enforced. In charismatic
parties, career advancement takes place through personal ties. Politicians’ careers
informal institutions and party organization 185
hinge on their personal relationship with the leader (Panebianco 1988, 145). Leader-
ship bodies may be reorganized at the leader’s discretion, and individuals without
careers in the party—often family members and cronies—may gain top positions. In
machine parties, career advancement takes place through patronage networks. Ac-
tivists are recruited into networks led by sitting or aspiring public officeholders and
ascend through the ranks when those politicians are elected to higher office.
Formal and informally organized parties differ with respect to their linkages to
social and civic organizations, such as labor confederations or business associations.
Where party organization is formal, these linkages are explicit and, in most cases,
written into party statutes. Formal linkages may take the form of party-sponsored
labor confederations, corporatist organs of representation within party leaderships
(such as the PRI’s labor and peasant sectors in Mexico, or the Labor Bureau in
Venezuela’s AD), or statutory women’s or youth wings. These linkages may be rein-
forced by formal rules such as quotas for candidacies and leadership posts, bloc votes
in party congresses, and collective peasant or union membership.
186 flavia freidenberg and steven levitsky
Where party organization is informal, societal linkages are not explicitly recog-
nized. Parties and social groups may routinely exchange human, financial, and
organizational resources, but these relationships are not institutionalized in formal
statutes or organizations. Rather, they take the form of loosely structured—often
patronage-based—alliances. These alliances may be with traditional social actors
such as unions or business associations, but they may also be with soccer fan clubs,
ethnic or religious groups, neighborhood associations, or nominally nonpartisan
civic organizations.
Finally, formal and informally organized parties differ in terms of finance. Where
party organization is formal, fundraising is done through legal and transparent chan-
nels. Both private and public contributions are regulated by the state. Private finance
is often subject to limits on contributions and rules requiring disclosure of all contri-
butions. Public finance is legal and transparent, often in the form of state sub-
ventions and free media time. In most formally organized parties, private and public
finance passes through the central bureaucracy (i.e., the party treasury), as opposed
to particular factions or politicians.
Where party organization is informal, finance is nontransparent, unregulated,
and often illicit. Parties do not keep reliable records of private contributions, and
rules of transparency and public disclosure either do not exist or are not enforced.
Private contributions routinely exceed legal limits, and they may include payments
from illicit sources, such as mafias, drug cartels, or foreign actors. Informally orga-
nized parties also make widespread use of unregulated—and often illicit—forms of
public finance. These include the widespread appropriation of public sector jobs for
patronage use, the use of kickbacks on government contracts to fill party coffers, and
the embezzlement of money and other resources (food, transportation) from state
agencies. In these cases, finance rarely passes through the central bureaucracy.
Rather, it is channeled through individual candidates and patronage networks.
These nine dimensions are summarized in table 9.1. Latin American party orga-
nizations are highly diverse (Alcántara Sáez and Freidenberg 2001; Alcántara Saez
2004) and may thus be found throughout the formal-informal spectrum. Some,
including the Brazilian PT, Mexican PAN, Venezuelan AD, and Chilean PCCh,
score fairly consistently on the formal end of these dimensions and thus can be
characterized as formally organized parties. Others, including the traditional clien-
telistic parties in Brazil, Colombia, Ecuador, Honduras, and Paraguay and person-
informal institutions and party organization 187
alistic parties in contemporary Peru, are predominantly informal. Many other par-
ties, including the Argentine Radical Civic Union, Peru’s Aprista Party, and Mexico’s
PRI, exhibit a mix of formal and informal characteristics.
To highlight how informal party structures work, this section examines two infor-
mally organized parties: the Argentine PJ and the Ecuadorian PRE. Although the
two parties differ in important ways (the PJ is patronage-based, whereas the PRE
mixes elements of personalism and charisma), in both cases, informal organization
predominates over the formal bureaucracy, producing behavior that deviates sub-
stantially from party statutes.
table 9.1
Continued
Formal organization Informal organization
mandate expired and the party congress failed to elect a new council, the body was
left vacant, leaving Peronism without any formal leadership whatsoever.
The PJ’s central bureaucracy is strikingly underdeveloped. The party lacks a
professional staff and has virtually no record of membership, finances, or activities.
Beyond custodial personnel, its headquarters are often empty. When Roberto García
became party president in 1990, the PJ headquarters were located ‘‘in a small office
that didn’t even have a sign outside.’’Ω Local party offices are equally underdevel-
oped. In 1997, the PJ’s Buenos Aires headquarters had no paid staff and was open
only twice a week, and the Tucumán headquarters lacked funds, phone service, or a
functioning bathroom for much of the year (Levitsky 2003b, 73). The bulk of Peron-
ist party activities are channeled through informal patronage networks, or agrupa-
ciones. These networks are not mentioned in the party statutes, and local party
offices generally keep no record of them. Yet they function as the de facto PJ organi-
zation, financing and coordinating local party activities, selecting candidates, and
organizing campaigns. As one local leader put it: ‘‘The party bureaucracy just main-
tains the headquarters, which is nothing more than an office and the ten employees
190 flavia freidenberg and steven levitsky
who run and clean the place. The rest of the organization is financed and coordi-
nated by the various agrupaciones. The party’s real infrastructure . . . is in the hands
of the agrupaciones.’’∞≠
Peronism’s grassroots infrastructure is almost entirely informal. On paper, the PJ’s
base-level organization takes the form of base units (UBs), or neighborhood-level
offices that fall under the jurisdiction of the local party council. In practice, however,
UBs are autonomous of the party bureaucracy. They are created and operated pri-
vately by Peronist activists. Anyone may open a UB anywhere, at any time. Fre-
quently, activists create UBs in their homes and literally become their ‘‘owners.’’
Local party bureaucracies thus have no influence over the number or location of
UBs. Indeed, most have no record of the UBs under their jurisdiction. Only a
minority of UBs take the form of standard party offices, with most taking the form of
informal ‘‘working groups’’ based out of activists’ homes or operating informally out
of unions, clubs, or community centers (Levitsky 2003b, 68–69). Although these
subunits do not formally exist (they are not recognized by the party bureaucracy),
they recruit members, compete in primaries, and campaign in elections.
The PJ’s boundaries are fluid and ill-defined. Peronism’s membership criteria
were never clear or widely enforced. Historically, individuals and groups simply
declared their affiliation. As a result, the party has at times contained individuals and
groups—including leftist and fascist paramilitary organizations—whose affiliation is
disputed. Indeed, the Montonero guerrillas made a point of not joining the party
during the 1970s. Although individuals or groups are sometimes expelled, these
expulsions are often ignored and are rarely permanent. Thus, Perón twice expelled
Catamarca party boss Vicente Saadi, but Saadi never lost control over Catamarca
Peronism, and although the Montoneros were ‘‘expelled’’ in 1974, most Montonero
networks never left Peronism. At the same time, individuals and factions routinely
float in and out of Peronism, abandoning the party for an election or two, only to
return a few years later. In 1985, for example, ‘‘Renovation’’ factions in several prov-
inces abandoned the PJ and competed against it in elections, but returned to take
over the party leadership a few months later. Due to these temporary schisms, it is not
unusual for two or more Peronist parties to compete in an election.
Peronist career paths are decidedly nonbureaucratic. During the party’s formative
years, careers hinged on individuals’ personal ties to Perón (Zorrilla 1983). Although
the PJ formally adopted a primary system in 1987, contemporary career paths are
shaped more by patronage networks than by elections. Peronists ascend through the
ranks via agrupaciones controlled by local and provincial bosses. Primaries are rou-
tinely undercut by ‘‘unity lists’’ imposed by party bosses (De Luca et al. 2002).∞∞ In the
absence of bureaucratic procedure, movement into, up, and out of the PJ hierarchy
informal institutions and party organization 191
is quite fluid. Juan Manuel Abal Medina (in 1972), Isabel Perón (1974), and José
María Vernet (1984) ascended to the top of the Peronist hierarchy without ever
holding a party post, and, between 1983 and 1990, none of the PJ’s four presidents
completed his mandate.
Justicialista Party membership is largely informal. During the 1990s, the party’s
membership—3.85 million, or 18 percent of the electorate—exceeded memberships
of most European social democratic parties (Levitsky 2003b, 61). However, PJ mem-
bership entails a lower level of commitment than does membership in many Euro-
pean parties. No application or screening process exists. Members do not pay dues
or attend regular meetings. Prospective members simply sign a form, and in the
absence of official oversight or enforcement, many registration forms are forged.
Given the lack of reliable membership rolls, many affiliates are not even aware of
their status.
Peronism maintains strong ties to a variety of working- and lower-class organiza-
tions, but these linkages are almost entirely informal. Historically, the PJ’s closest ties
were to unions (Torre 1990; McGuire 1997). However, unlike most European social
democratic parties, these linkages were never formalized. Traditionally, the party-
union linkage was based on two informal mechanisms: the ‘‘62 Organizations’’ (or
‘‘62’’) and the tercio system. The ‘‘62,’’ which functioned as the PJ’s ‘‘labor branch,’’
nominating unionists for candidacies and leadership posts, was not mentioned in
Peronist statutes and had no formal position in the party leadership (McGuire 1997,
98–99). Similarly, the tercio tradition of granting unions a third of candidacies and
leadership posts was not found in the statutes. The PJ also maintains informal
linkages to a variety of other working- and lower-class organizations, including squat-
ters’ organizations, neighborhood associations, clubs, church groups, and a variety of
other civic and nonprofit groups (Levitsky 2003b, 62–65). Of particular importance
are ties to local soccer clubs. Soccer clubs are a central part of social life in working-
class zones, and links to these clubs allow the PJ to tap into organized fan clubs to
mobilize youth for electoral campaigns. Soccer fan clubs are mobilized to attend
rallies, paint campaign graffiti, and, in some cases, engage in thuggish activities
aimed at rival factions or parties.
Finally, the PJ is financed largely through informal channels. On paper, the PJ is
financed through a combination of state subsidies and private contributions, which
are channeled through the party bureaucracy. However, these regulated forms of
finance are dwarfed by unregulated contributions. In 1989, for example, the PJ
officially reported taking in a total of US$1.8 million in contributions, but one
businessman claimed to have given US$3 million to the Menem campaign (Ferreira
Rubio 1997, 19). The most important sources of party finance, however, are pa-
192 flavia freidenberg and steven levitsky
port, launched a bid for the post of supreme director without Bucaram’s blessing
during the 1997 National Convention, Bucaram declared that the position was no
longer up for contestation and that party authorities would remain in their posts for
another year (Freidenberg 2003, 171). This response openly violated party statutes,
which called for an annual renovation of the leadership.
The PRE’s central bureaucracy is weak and ineffective. According to its 1982
statutes, the PRE is organized in a bureaucratic manner: national, provincial, and
local commands integrate the party’s membership and diverse internal groups into a
formal hierarchical relationship. In practice, however, this structure is ignored. The
central bureaucracy is an empty shell. Local and provincial commands exist only in
the few districts where the PRE is particularly strong (Freidenberg 2003). Most local
party offices are not integrated into the bureaucracy, but rather are privately owned
or rented. Party subunits are often located in local bosses’ homes or businesses. Each
local leader possesses the autonomy to create offices where he wishes, and to finance
and manage them as his own. The PRE’s national headquarters in Guayaquil has no
record of these local branches and thus cannot coordinate or control them.
The PRE’s local infrastructure is based almost entirely on personal and patronage
networks. Local bosses build personal followings—through family, commercial, or
sports ties—at the margins of the party bureaucracy (Freidenberg 2002). Particularly
in the PRE’s electoral strongholds along the coast, these networks are critical to the
party’s electoral success: they deliver votes, money, local media influence, contacts
with the local elite, and, crucially, groups of supporters who carry out party work.
Local bosses enjoy substantial autonomy in organizing the party in each district,
although this autonomy is ultimately limited by their subordination to Bucaram.
Career paths in the PRE also diverge from bureaucratic norms. Advancement
within the party depends largely on personal relationships, particularly to Bucaram
and his family (Freidenberg 2003, 198–206). Most top party officials are family mem-
bers, longtime cronies, or business partners of Bucaram. For example, Bucaram’s
wife has served as PRE supreme director (the only person other than Bucaram to
hold this post); his brother Adolfo has served as national subdirector during his exile;
another brother, Jacobo, was the PRE presidential candidate in 2002; his sister Elsa is
a member of the National Command and has been PRE mayor of Guayaquil; and
his eldest son, Jacobito, is the de facto leader of the PRE’s parliamentary bloc—even
though he is not even a member of Congress. Several of Bucaram’s children, neph-
ews and nieces, cousins, and in-laws also occupy positions in the party hierarchy.
Bucaram’s longtime business partner Eduardo Azar Mejía is responsible (along with
Bucaram and his brothers) for the PRE’s finances.
Notwithstanding the existence of party caucuses and other formal leadership and
194 flavia freidenberg and steven levitsky
candidate selection processes, Bucaram and his family dominate the hierarchy and
are usually decisive in shaping career paths. The PRE’s candidate lists are routinely
altered—usually at Bucaram’s request—after they have been selected by party leaders
(Freidenberg 2003). Party bosses play a similar role at the local and provincial levels.
According to Bucaram, ‘‘the provincial commands know nothing about internal
elections. They run them, but they don’t care about them. It is always the director,
the provincial leader, who chooses the names’’ for candidacies and party leader-
ship posts.∞∂
The PRE’s organizational boundaries are ambiguous. More than a party, mem-
bers belong to the Roldosista ‘‘family,’’ which is viewed as broader than the PRE
itself. The party lacks clear criteria for membership or updated membership rolls. As
a result, membership in the PRE is largely a matter of self-declaration. Joining the
party—at virtually any level—is simple, especially if one has financial resources, a
clientelistic base, media access, or ties to the supreme leader (Freidenberg 2001,
2002). Indeed, local caciques (‘‘political brokers’’) move with relative ease between
the PRE and other clientelistic parties. Unlike Argentina’s PJ, affiliations and expul-
sions are subject to the ultimate approval of Bucaram, but this vetting process—even
if later ratified by the party’s National Convention—is thoroughly informal.
The PRE has no formal membership requirements or selection criteria, collects
no dues, and possesses no record of its own membership. Activists are generally
recruited through clientelistic networks, joining the party in the hope of gaining
access to material benefits. Particularly in low-income zones, many party members
view the Roldosista family as a salvation in a context of social anomie and scarcity of
resources. Membership may entail a relatively high level of commitment. To gain
access to material benefits, new members must prove themselves to local party bosses
by carrying out a variety of organizational tasks during electoral campaigns. There is
also a charismatic element to party-activist linkages: party loyalties are reinforced by
followers’ belief in the extraordinary qualities of Bucaram’s leadership.∞∑
The PRE maintains strong informal ties to diverse social organizations in low-
income regions, particularly along the coastal region of Guayaquil. These include
food kitchens, human rights groups, and associations of small retailers, banana
growers, lawyers, and social workers. Though not formally linked to the PRE, many
of these organizations carry out intensive party work. In addition, many PRE bosses
either directly own or maintain close ties to soccer clubs, using members of soccer
fan clubs as advisors, bodyguards, and rank-and-file activists. Soccer players them-
selves often work in PRE offices, campaign for the PRE, and serve as bodyguards for
PRE legislators.
informal institutions and party organization 195
The PRE finances itself through both formal and informal channels. Like all
Ecuadorian parties, the PRE receives financing from the state’s Permanent Party
Fund. Although party statutes stipulate that these monies be channeled into the
party bureaucracy via an economic commission in the National Command, party
leaders acknowledge that no such commission exists. It is the supreme leader and his
close associates—including private business partners—who control the party’s fi-
nances and coordinate the bulk of its fundraising. Although it is impossible to
measure accurately the unregulated finance received by the PRE, it is clear that this
quantity exceeds that of regulated finance. The PRE makes massive use of illicit state
finance, including government jobs, food, medicine, public housing access, and a
variety of social programs to reward activists and voters (Freidenberg 2001). State
resources are complemented by the private contributions of local bosses. In many
cases, it is the local bosses themselves (and not the party) who pay for local advertis-
ing and finance transportation for activists, hire graffiti-painting brigades, and pro-
vide food and live music for party rallies.
Like Argentina’s PJ, then, the PRE’s formal statutes tell us little about how it
makes decisions, selects its leaders and candidates, organizes its activists, or raises
money. The major difference between the PRE and the PJ is that the former com-
bines an informal structure with a charismatic leadership. Whereas contemporary
Peronism is organized along patronage lines, Bucaram’s charismatic leadership—
like Perón’s in the original Peronist Party—is critical to the PRE’s electoral mobiliza-
tion, its activist base, and its linkages to society.
conclusion
This chapter has sought to (re)introduce the dimension of informality into con-
temporary analyses of party organization. Although this dimension has long been
central to research on political clientelism and machine politics, informal institu-
tions have been confined to the margins of contemporary conceptual and theoretical
work on party organization.
Informal organization matters because it shapes how parties work. Thus, a crucial
area for future research lies in exploring ways in which informality affects party be-
havior. For example, recent studies suggest that informally structured parties may be
more flexible and adaptable in the face of environmental change than are bureau-
cratic parties (Roberts 1998; Levitsky 2003b). Informality also affects candidate selec-
tion. Studies of parties in the United States suggest that primary systems may en-
courage the nomination of ideologically polarizing candidates, as turnout tends to
196 flavia freidenberg and steven levitsky
be higher among ideologically committed voters (Wright and Berkman 1986). Yet
in clientelistic or patronage-based parties, in which local brokers bring voters to
the polls, primaries produce very different results: elections are almost always won
by candidates backed by the largest patronage organization or machine. Similar
questions may be asked about legislative cohesion, coalition formation, activist re-
cruitment, intraparty democracy, and public policymaking in informally organized
parties.
Another potential area for research is how informal party organization affects the
quality of democracy. At first blush, these effects seem to be uniformly negative.
Informal party structures are often rooted in particularistic relationships, which tend
to erode the kinds of programmatic linkages that many scholars view as critical to
effective democratic representation. In addition, the nontransparent nature of most
informal party structures limits democratic accountability. Where decision-making
power lies outside formal authority structures, formal mechanisms of democracy
and accountability—such as party caucuses, congresses, and primaries—tend to be
undercut. (This is clearly seen in Langston’s chapter, in this volume, on the PRI.)
Where finance is unregulated and undocumented, and parties’ real power brokers
thus obscured from public view, voters’ capacity to discern who and what those
parties represent is limited. Hence, informally organized parties may be less oriented
toward public goods provision and more vulnerable to capture by private—even
illicit—interests.
Informality may not always be inimical to democracy, however. In the context of
repressive or exclusionary regimes, informal party structures may be critical to the
survival or mobilization of disadvantaged groups. For example, Peronism’s informal
structure helped it survive two decades of proscription and several periods of military
repression, and informal networks helped sustain resistance to Polish communism,
South African apartheid, martial law in Taiwan, and various other dictatorships.
Informal organization may also provide channels of access to groups that have
traditionally been excluded from the political process. In Ecuador, for example, the
populist Concentration of Popular Forces incorporated previously marginal sectors
during the 1960s (Menéndez Carrión 1986), and Pachakutic—which mixes elements
of formality and informality—vastly expanded indigenous peoples’ electoral and
legislative representation during the 1990s.
Latin American parties are now being studied with increasing comparative
breadth and theoretical sophistication. Theories that for years were limited to studies
of advanced industrialized countries are now being applied and tested in Latin
American cases. Although these are indeed positive developments, efforts to theorize
Latin American parties must be based on an accurate understanding of how those
informal institutions and party organization 197
parties work on the ground. The gap between how parties are organized on paper
and how they function in practice is much wider in Latin America than in most
advanced industrialized countries. Bridging that gap requires that scholars move
beyond party statutes and enter the murky world of informal politics—a world than
in many cases revolves around patron-client ties, corruption, and other less-than-
transparent activities. This is difficult empirical terrain. Yet as long as scholars steer
clear of it, they will be left with an incomplete (and inaccurate) picture of how
political parties work.
This page intentionally left blank
part iv
daniel m. brinks
Democracy, most students of the region agree, has swept across Latin America
since the 1970s and early 1980s. But most also agree that democracy in the region
suffers significant shortcomings, and one of the shortcomings most often cited is the
failure of the rule of law. Thus, Michael Shifter points to two key components of a
democratic rule of law: ‘‘Latin American Democracy is most seriously stalled on two
key fronts. The first is a drive for a legal system that guarantees both the equality of all
citizens before the law and basic personal rights. The second has to do with the
separation of powers and the imposition of effective checks on executive authority’’
(1997, 116). One of the critical components of what Guillermo O’Donnell has called
the ‘‘brown areas’’ of democracy is precisely this failure. O’Donnell argues that in
most countries of Latin America, ‘‘huge gaps exist, both across their territories and in
relation to various social categories, in the effectiveness of whatever we may agree
that the rule of law means’’ (1999c, 311).
This chapter examines one area in which the legal systems of Brazil and Argen-
tina have failed to check executive authority (in the sense that the police are agents
of the executive) and failed to guarantee basic personal rights: the police in both
Brazil and Argentina, with some variation across and within the countries, continue
to kill with impunity, despite laws that purport to limit their use of lethal force. In
1992, for example, the police in São Paulo killed more than fourteen hundred
people. Over the course of the 1990s, this police force averaged nearly seven hun-
dred homicides per year. While these absolute numbers are virtually unparalleled in
the region, per capita, the police in Salvador da Bahia (in northeastern Brazil) kill
202 daniel m. brinks
three times as often, while the police in Buenos Aires kill almost as often as those in
São Paulo. Meanwhile, conviction rates for police officers accused of homicide
hover around 7 percent in São Paulo, less than 5 percent in Salvador da Bahia, and
below 20 percent in Buenos Aires.
The main goal of this chapter, then, is to examine the mechanisms and processes
that give rise to the high levels of impunity for police officers who kill in the course of
their duties. One of the arguments that have been put forth to account for the failure
of the rule of law generally (as well as the failure of various other institutions to
produce their intended results) is the presence of informal institutions. From the
pattern described above, one might conclude—and many human rights activists in
the region argue—that there is an informal institution that gives the police broad
latitude in the use of lethal force: the police may kill whom they must, so long as they
maintain some minimum of social stability. The hypothesized informal institution,
therefore, is one that grants the police broad or even unlimited discretion in the use
of lethal force.
If (formal) institutions are made up of (formal) rules, then informal institutions
must also be made up of (informal) rules. Thus, if informal institutions are driving
police impunity, we should be able to detect informal rules that, at a minimum,
permit this behavior. With due apologies to Méndez, O’Donnell, and Pinheiro
(1999), then, the presence of informal institutions might be labeled ‘‘the rule of
(non)law.’’ Rules that are not laws determine outcomes. But before we can evalu-
ate this argument, we need a more precise definition of informal institutions (and
perhaps even of rules) and a more precise notion of how to operationalize these
concepts.
The second goal of this chapter, therefore, is to specify and apply a rigorous
definition of informal rules and informal institutions, and to propose some standards
for operationalizing informal institutions. This second goal necessarily must be dealt
with first, so I begin by expanding upon some of the theoretical developments already
anticipated in Helmke and Levitsky’s introduction to this volume, and then apply this
theoretical framework to the institution at issue here. I conclude by examining an
instance of institutional change affecting Salvador da Bahia and placing the institu-
tion at issue within the classification scheme proposed in the introduction.
The introduction to this volume and an earlier essay (Brinks 2003b) develop
much of the definition of informal institutions. I will not repeat that discussion here,
but will sketch out the basic elements of this definition and, along the way, propose a
prosecuting police killings in brazil and argentina 203
series of questions that can guide researchers in determining whether certain be-
havior responds to an informal institution.
As noted in Helmke and Levitsky’s survey of the concept, the study of informal
institutions generally requires a more ethnographic approach and needs to be more
data intensive than a survey of formal institutions. The first two questions posed here
will permit researchers to quickly determine whether we need bother to look for an
informal institution at all.
If the answer to this question is no, we need go no further. Informal rules derive their
existence in part from the very fact of their operation (Brinks 2003b). While laws and
other formal rules may, in some sense, continue to be laws even if they are never
enforced, it makes little sense to say there is an informal rule but it is never applied.
And it makes little sense to waste research resources on a hypothetical institution that
ultimately makes no difference. We may come to this research from a different and
equally valid starting point, however, if we find that relevant actors consistently
ascribe certain outcomes to informal rules.
Depending on the rule, of course, informants may not be willing to acknowledge its
existence. On the other hand, informants are often more than willing to describe
widespread illegal practices as the norm. In either event, they may be completely
sincere—and completely mistaken. As a result, while the description of an informal
rule can form the starting point for the inquiry, we still need to find more objective
evidence of its existence.∞
Once we are on the lookout for an informal institution, however, we need to be
precise about what we mean by the concept. There is broad agreement that informal
institutions, like formal ones, are made up of rules. Helmke and Levitsky define
informal institutions, in contrast to formal ones, as ‘‘socially shared rules, usually
unwritten, that are created, communicated, and enforced outside officially sanctioned
channels.’’ In an earlier essay (Brinks 2003b), my distinction between formal and
informal institutions focused purely on the way in which informal rules are created,
not on the means of communication and enforcement. I argued that informal rules
are those that are not enacted in accordance with the second-order rules that govern
the creation of formal rules in a given organizational context (which is, I think,
204 daniel m. brinks
the relevant agents of social control.≥ This leads to the crucial question for identify-
ing such an institution.
Question 3: Are deviations from the hypothesized informal rule punished by the
relevant agents of social control?
This evidence, where available, is the acid test for the presence of an informal
institution. As Ellickson (1991) notes, people might regularly behave in a certain
manner for any number of reasons, unrelated to the presence or absence of rules
governing that behavior. The very best evidence of the existence of a rule is the
punishment of deviant behavior by third-party agents of social control (Hart 1961, 9–
25; Ellickson 1991, 128). These agents could be neighbors, bosses, the police, the
courts, or many others, depending on the type of institution at issue.
This apparently simple matter requires a little more elaboration, however. Super-
ficially, rules can take three forms: they can permit, require, or prohibit certain
behavior (Crawford and Ostrom 1995, 585). The latter two are self-evidently equiva-
lent: a rule that requires x prohibits not-x. In either case, the search for deviant
behavior is quite simple, as the deviant actor is the one engaged in the primary
behavior that the rule addresses. But a permissive rule generates a different kind of
deviant behavior. The actor to whom the rule is apparently addressed cannot trans-
gress, as both x and not-x are permissible. In this chapter, for example, I hypothesize
a rule that permits but does not require the killing of violent criminals. In a given
incident, a police officer has a choice whether or not to kill and, I argue, will not be
punished either way. If this is true, the decision to kill can be made without factoring
in the cost of a sanction for violating the rules against homicide.
A rule that permits either x or not-x is in reality directed at agents of control and
prohibits punishing either of these choices. It could be restated as: ‘‘You may not
punish agent A for doing either x or not-x.’’ Deviant behavior for such a rule, there-
fore, is the attempt to punish behavior that falls within the protection of the informal
rule. The best evidence of a permissive rule is proof that someone in a position of
authority tried to enforce a contrary rule and was punished for doing so. In the case I
am considering here, then, the best evidence of a permissive rule regarding police
homicides would be finding that a prosecutor or judge was punished for attempting
to prosecute a police officer who exceeded formal limits on the use of lethal force,
but was acting within the hypothesized informal rule.
But the effectiveness of a rule may well be measured precisely by the infrequency
with which it is violated, so that, paradoxically, the strongest rules will show the least
evidence of enforcement—at least once they are well established. If we cannot
Beginning step: Outcomes do not match formal institutional predictions
and/or
People describe certain behavior in terms of an enforceable rule of conduct
䉲
Key question: Are violations of the hypothesized informal rule punished by relevant agents of social control?
—If rule requires or prohibits certain behavior, look to primary behavior for violations
—If rule merely permits certain behavior, look to secondary behavior for violations
䉲 䉲
If yes: This is primary evidence that an If transgressions are not observable, or as
informal institution is at work additional proof, look for secondary evidence:
䉲 䉲
Do relevant actors: 1) know the If the rule permits illegal behavior,
rule, and 2) act in anticipation is the behavior observed and not
of enforcement action? punished?
No negative sanctions should be applied to those who act under the aegis of the
informal rule, even if it permits or requires formally illegal conduct. Here, we must
be careful to distinguish between weak or nonexistent enforcement of a rule and
actual tolerance of violations by those in a position of authority. Assuming that
enforcement agents are at least minimally capable of exercising their function, the
key will lie in determining whether these agents observed the behavior in question,
or if it simply went undetected. Failing this, or as additional proof, we might resort to
indirect evidence of the existence of an informal institution.
Question 5: Is there evidence that relevant actors know the rule, anticipate the
consequences of a transgression, and guide their conduct accordingly?
One of the functions of a rule is precisely to use the threat of sanctions to prevent
repeated violations. Thus, if actors acknowledge that their behavior is constrained by
the anticipated secondary consequences of failing to follow the rule, we may con-
clude that a rule is at work. Here, in essence, we are looking for direct testimonial or
other evidence of the shared normative expectations called for in the definition of
rules. In my case, in which the proposed rule permits otherwise illegal conduct, we
should see the actors in question not only engaging in this conduct, but doing so very
openly, taking little or no pains to hide their illegal conduct from formal enforce-
ment agents.
All these questions can be organized into a decision tree for the identification of
informal institutions that permit otherwise illegal behavior (figure 10.1).
In the next section, I apply these questions to an observable behavioral regularity:
the use of lethal force by the police in five South American cities: São Paulo and
Salvador da Bahia in Brazil, Buenos Aires and Córdoba in Argentina, and Monte-
208 daniel m. brinks
video in Uruguay. I conclude that on the one hand, there is no evidence to suggest
that an informal institution protects this practice in Uruguay and in Córdoba, while
at the other extreme, in Salvador da Bahia, there is broad permission for police
killings of marginal social characters. Between these extremes, in Buenos Aires and
São Paulo, the rule permits the killing of particularly violent criminals.
As noted, the first step is to identify a regularity that is not attributable to a formal
institution, a structural condition, or other identifiable feature of the sociopolitical
landscape. The second step is to determine whether the observed regularity con-
stitutes an informal institution. In order to make this determination I go through the
five questions presented above for each of the locations in the study.
The most visible behavior prompting this research is the regular and illegal use of
lethal force by the police, despite laws that purport to outlaw the practice. And the
hypothesis proposed to explain this behavior is the existence of an informal institu-
tion that permits the indiscriminate use of lethal force by the police in the various
countries of Latin America. As explained above, however, because the rule is a
permissive one, the actual conduct that the hypothesized rule purports to proscribe
is punishment on the part of agents of social control (e.g., judges and prosecutors) of
police officers who act under the protection of the rule. I take the first behavior—the
police killings themselves—as a given, without seeking to explain it, in order to ex-
plore the second aspect—the failure of enforcement mechanisms to punish murders
committed by the police.∂
Clearly, the police conduct at issue is inconsistent with the formal rules. The
substantive laws on this matter are essentially the same in Argentina and Brazil. Both
countries have signed international conventions that protect human rights, includ-
ing the right to due process, the right to access the legal system, and the right to be
free from summary execution. Both countries have incorporated these laws into the
domestic legal framework. Recent constitutional changes in both countries have
elevated basic civil rights and human rights treaties to constitutional status, includ-
ing protections against arbitrary police violence. Uruguay is also a signatory to the
main international human rights treaties and has the standard protections in its laws
against arbitrary killing, though it has done less in the postdictatorial period to
strengthen its domestic laws in this area than either Argentina or Brazil.
prosecuting police killings in brazil and argentina 209
Despite these laws, in the last decade, the number of civilians killed in São Paulo
and Buenos Aires has been very high and shows signs of worsening. On average, the
police in the Buenos Aires metropolitan area killed 160 persons every year from 1990
through 2000. The worst year was 1999, when they killed 277 persons—more than 5
people every week. The trend has been generally rising. In São Paulo the numbers
are even higher, averaging 680 victims annually from 1990 through 2000. The num-
ber peaked in 1992, at a stunning 1,428 persons. Even considering the immense
population of the state of São Paulo—around thirty-six million—this number of
casualties is reminiscent of a civil war. The difference in the size of the two cities
obscures the similarity between per capita figures for Buenos Aires and São Paulo.
From 1990 to 1994, São Paulo’s police homicide rate was twice as high as that of
Buenos Aires: 2.2 deaths per 100,000 versus 1 per 100,000. But from 1995 through
2000, the rates are nearly identical: 1.61 and 1.63 per 100,000.∑
By way of contrast, other cities with large populations and police forces and with
notorious violent crime problems show indices that are orders of magnitude lower in
both absolute and relative terms. In 1992, São Paulo police killed thirty-one people
for every one shot to death in Los Angeles, the city whose police ‘‘kill more people in
proportion to the size of the force than any other major U.S. police department’’
(Chevigny 1995, 46). Los Angeles reported approximately 0.5 killings per 100,000 in
1991 and 1992. New York’s figures are lower: an average of about 0.34 per 100,000
persons killed per year between 1991 and 1993. The best available number for a single
year for Mexico City, an urban area larger even than São Paulo, is less than 1 per
100,000, though this is likely due, at least in part, to undercounting (Chevigny 1995).
Complete information on police homicides in Salvador da Bahia is hard to come
by. There is no official oversight agency other than the regular prosecutor’s office,
and the relevant nongovernmental organizations (NGOs) do not regularly collect
systematic information on the subject. But what we know about police homicides in
Salvador suggests that the incidence of violations is even higher, and that judicial
results are possibly worse, than in São Paulo. A study by the Justice and Peace
Commission of the Archdiocese of Salvador, based on a survey of daily press reports,
counted 623 deaths at the hands of police officers in the Salvador metropolitan area
between January 1996 and December 1999 (de Oliveira et al. 2000). Given a popula-
tion of 2.5 million, the average annual rate of deaths per 100,000 for these four years
in this metropolitan area was 6.23, nearly three times higher than São Paulo’s average
for the most violent half of the last decade.∏
In Uruguay and in Córdoba, Argentina, the situation is quite different. A search
of online archives for the main newspapers in Uruguay, plus a review of the human
rights reports by international NGOs and the police violence reports prepared by the
210 daniel m. brinks
Peace and Justice Service, a Uruguayan NGO, reveal only a little more than twenty
killings from 1992 to 2000 in all of Uruguay. In Córdoba, the Council against Police
and Institutional Repression, an NGO that is devoted exclusively to killings com-
mitted by state agents, reports an average of about ten cases per year over the course
of the 1990s.
The legal response to this activity strongly suggests that, at least in some instances,
the courts are prepared to tolerate this level of violence. In Salvador da Bahia, São
Paulo, and Buenos Aires, the conviction rate is so low that it can be readily construed
as official complicity in the conduct in question.π Figure 10.2 compares the propor-
tion of convictions for all five locations, based on a database I compiled of more than
five hundred prosecutions of police homicides in São Paulo, Buenos Aires, Salvador
da Bahia, Córdoba, and Uruguay, to the per capita rates of police homicides.∫ The
conviction rates include only cases that involved official police conduct, whether in
or out of uniform, and exclude cases in which police officers killed someone in a
quarrel or similar private dispute.
Note that the locations with the highest rates of police homicides have the lowest
conviction rates. If the formal rules were the only ones at work and the enforcement
capacities of all these systems were even remotely similar, we might have expected
quite the opposite: the more restrained police forces should be using force only
under more justified circumstances, thus producing a lower conviction rate, while
the more violent forces in Salvador da Bahia, São Paulo, and Buenos Aires should be
punished more often. The answer to the first question, then, is yes, there is a be-
havioral regularity—the failure to convict police officers who kill—that cannot be
explained by differences in the formal rules of the game, in at least three of the cities
at issue.
But the exact contours of the rule are not so clear. On the one hand, there seems
to be little reason to believe that a permissive rule is present at all in Córdoba and
Uruguay. Conviction rates there are quite high, despite what seem to be less violent
police forces. Moreover, not all of these cases go completely unpunished in even the
worst systems—this is more true in Buenos Aires, but is even true in São Paulo, where
6 percent or so of the cases end in a conviction. Thus a blanket permissive rule is
somewhat inconsistent with the evidence. Based on comments made by various
activists whom I interviewed in May through December 2000 and my own observa-
tion of a few prominent cases in Buenos Aires that could not be otherwise explained,
the rule might work only in cases in which the victim had forfeited the right to legal
protection by taking up arms against society.
I set out to test for the existence of this informal rule of decision, using the
database described earlier. I coded all the cases to identify those that would fit within
prosecuting police killings in brazil and argentina 211
Figure 10.2. Police Homicides and Conviction Rates in the 1990s. Sources: Data from multiple
official and unofficial sources; see note 8.
the parameters of the rule, and compared judicial outcomes across the categories
defined by the rule. The hypothesized rule, broadly speaking, frees the police to kill
individuals perceived as being violent criminals, whether they shoot them in the
course of a gun battle or execute them in cold blood after they have been ap-
prehended. For ease of reference, I call this the Violent Victim exception.
In practice, of course, victims may be more or less tainted by a connection with
violence, and the evidence of that involvement may be more or less firm, so the rule
works more as a sliding scale, with the level of prosecutorial and judicial resources
devoted to a case going down, producing a lower conviction rate, as the cases ap-
proach the extreme. For purposes of this discussion, however, I included only clear-
cut cases in this category, coding as Violent Victim cases those in which there was
strong third-party evidence—in the form of eyewitness reports, criminal records, or
detailed news coverage—that the victim had a close connection to a violent crime,
whether or not the killing occurred in connection with that crime. Table 10.1 com-
pares results in these cases with all other cases in the database, for all sites except
Salvador da Bahia.Ω
In São Paulo and Buenos Aires, none of the cases involving Violent Victims led to
a conviction. In Uruguay and Córdoba, on the other hand, there is no evidence that
this rule drives the outcomes. The Uruguayan system actually shows a higher convic-
tion rate in Violent Victim cases, but because of the small number of cases, the
rather large percentage difference across the two groups is not statistically significant.
In Córdoba there are simply not enough of these cases to ground a conclusion one
212 daniel m. brinks
table 10.1
Conviction Rates in Violent Victim cases
Conviction rate (total no.)
Non-Violent Violent Victim
City Victim cases cases All cases
way or the other. Two facts suggest that the rule does not operate in this city: first, one
of only two cases in this category ended in a conviction; and second, if the police felt
they had license to kill, we might expect more of these cases. Thus, in São Paulo and
Buenos Aires at least, there is a regularity that is more precisely explained by the
proposed informal rule: security agents who kill Violent Victims are not punished,
even though the law suggests they ought to be.∞≠
Moreover, I coded all the cases for evidence that the killing was an execution in
cold blood—that is, the victim was shot and killed after being reduced to custody and
control. In legal terms, these are extreme cases and should result more often, if not
always, in a conviction. As we might expect, there is a higher percentage of execu-
tions in the Violent Victim cases than in other cases: in both São Paulo and Buenos
Aires, 37 percent of all Violent Victims were executed, compared with 25 percent of
all others. But even in these extreme cases there are no convictions. In short, the
formal laws cannot account for this result, nor can it be explained by the strength of
the enforcement agencies, as the difference is repeated within each system. This
striking pattern, so neatly predicted by the informal rule and so difficult to explain by
reference to other variables, is already evidence in support of finding an informal
institution at work.
The pattern of impunity seems much broader in Salvador da Bahia. A study by
Lemos-Nelson (2001) shows that members of the civil police who kill a suspect are
rarely investigated by internal disciplinary instances. Lemos-Nelson finds that in
most cases the very first step in a judicial investigation, the inquérito, is never com-
pleted, so the case does not even reach the prosecutor’s office or the judiciary. The
activists who spoke to me all agreed that only in the very rarest cases—cases that for
one reason or another served a political purpose—was an effective judicial response
likely. Moreover, I selected about thirty cases that received media attention in the
prosecuting police killings in brazil and argentina 213
second half of the 1990s and traced the judicial response to them. With the exception
of recent cases involving minors (addressed in more detail later in the chapter), I
found no convictions at all among these cases. I will be describing many of these
cases in more detail.
In addition to these regularities, as I have suggested already, there is abundant
testimonial evidence that an informal rule of some kind is operative in this regard,
which brings us to the next question.
In the course of fieldwork in Argentina and Brazil, I often came across statements
suggesting that the killing of a civilian by a police officer would be punished only if
the victim were somehow worthy of legal protection, independent of the facts sur-
rounding the killing. In fact, it is not uncommon to find statements in which victims’
relatives, the police, or even elected officials seem to approve of a policy of killing
those who constitute a danger to society.
Two different prosecutors in São Paulo whom I interviewed argued, for example,
that policemen who killed criminals would not be convicted in the courts, no matter
what the legalities of the case. Within the police corporation, police officers who kill
(and are assumed to have killed violent criminals) are often promoted—Gilson
Lopes in São Paulo, for example, consistently climbed the ranks of the military
police, in no small part due to his role in at least forty-two killings over the course of a
decade, before finally being convicted of homicide and removed from the force
(Chevigny 1995, 167). Carlos Ruckauf, when he was governor of the province of
Buenos Aires, became a notorious advocate of a meta bala approach to crime control
(roughly translated, a ‘‘shoot ’em up’’ approach). He offered a bonus of up to five
times their salary to police officers involved in ‘‘acts of bravery,’’ which he loosely
defined as the detention or killing of violent criminals.∞∞ This gesture echoes the
actions of the commander of the São Paulo military police, who also offered a reward
and a five-day leave to those who distinguished themselves by killing criminals
(Chevigny 1995).
Even the relatives of victims show that they understand public acceptance of the
use of violence against violent criminals. They always emphasize that their relative
was not a marginal. It is not uncommon to hear them justifying the killing of
‘‘lawless’’ criminals, even while protesting the unjustified murder of their own rela-
tive. A mother decrying the killing of her son, whose story is related by de Oliveira
and coauthors, says: ‘‘Why didn’t they go after misguided children? Because if they
214 daniel m. brinks
went after misguided children, and took their lives, they wouldn’t cause such hard-
ship, but they took a life that had nothing to do with crime!’’ (de Oliveira et al. 2000,
142). I heard echoes of this complaint in Salvador da Bahia, São Paulo, and Buenos
Aires, but not in Uruguay. In Córdoba, the issue was presented more as a matter of
class than criminal records; and indeed, in some convictions the victim had an
extensive record.
In Salvador, there is other testimonial evidence describing the rule. In a dossier
prepared by a group of human rights entities, the investigator in charge of the
homicide division in that city makes the point explicitly. He relates that whenever he
arrives at a police outpost to investigate the killing of someone who had a criminal
record, ‘‘they say ‘we have better things to do than to clear up the death of a crimi-
nal.’ ’’ Leo Ornéllas, the head of an Afro-Brazilian civil rights group, agrees: ‘‘In
Bahia, before investigating, [police investigators] first check to see if the deceased
victim had a criminal record, and then they decide whether or not to investigate.’’ A
priest who works in one of the most violent neighborhoods in Salvador relates similar
responses by police investigators. In one case, he says, the delegado refused to in-
vestigate the shooting of a young girl, simply because the girl’s mother was involved
with drugs.∞≤
The unidentified killers who execute marginais are known as justiceiros, a word
that can be translated as ‘‘vigilante,’’ but carries an obvious connotation of meting
out ‘‘justice.’’ In interviews, people, especially those of lower socioeconomic stand-
ing, often say ‘‘marginal tem que morrer mesmo’’—criminals must simply die (de
Oliveira et al. 2000, 35). Even the media are to some degree complicit in justifying
these killings. Commenting on the newspapers’ persistent habit of noting the crimi-
nal history of victims (even if unrelated to the incident itself ), the Bahian sociologist
Carlos Espinheira says: ‘‘It is as if there is a tacit understanding that a habitual
criminal can be eliminated, while one who has never been arrested was, in fact,
murdered, with the full moral and legal import society assigns to such an event’’ (de
Oliveira et al. 2000, 35; my translation). Thus several sources would suggest there is
an informal exception to the rule against the killing of civilians by the police: the rule
would permit the killing of criminals, though not, presumably, of ‘‘honest citizens.’’
It is at least a plausible hypothesis, therefore, that policemen who kill in the
course of their duties are protected by a broad and generous rule against their
prosecution in Salvador da Bahia, and that those who kill Violent Victims are
protected by a similar rule in Buenos Aires and São Paulo. But we need to test this
hypothesis with more than simply the presence of a behavioral regularity and some
testimonial evidence.
prosecuting police killings in brazil and argentina 215
The third question brings us to the enforcement activity surrounding the rule. As
noted earlier, because the rule in question here is permissive, we must look for
deviance and enforcement at the secondary level. What we are looking for is not
merely a lack of enforcement that is consistent with the informal rule and not with
the law, but, more importantly, the punishment of deviations from that pattern. This
information, as we will see, is not always easy to come by. Precisely because actors
anticipate and avoid negative consequences, we should expect these instances to be
rare. The absolute lack of convictions among Violent Victim cases in São Paulo and
Buenos Aires makes it difficult to determine what adverse consequences might flow
from an attempt to strictly enforce the law. One lawyer in Buenos Aires suggested
that a particular judge, whom she viewed as sympathetic to the victims’ cause, was
repeatedly passed over for advancement. While in both Buenos Aires and São Paulo
appellate courts intervened occasionally in favor of the defendants (and never in
favor of the victim) in these cases, the number of appellate interventions is too low to
support a firm conclusion on this issue.
I did obtain from one prosecutor and several other lawyers in Buenos Aires a
suggestion of how the rule might be enforced from below. They argued that if the
police viewed certain prosecutors or judges as being overly aggressive about enforc-
ing the law in cases in which the police felt they were simply doing their job, they
would simply stop cooperating with them. In Buenos Aires the courts depend on the
regular police for all the everyday tasks associated with the investigation and prosecu-
tion of individual cases.∞≥ Judges and prosecutors rely on the police to serve warrants,
locate witnesses and bring them into court, and serve the hundreds of oficios—a sort
of general-purpose writ requesting that something be done—that arise in the ordi-
nary course of business. When judges or prosecutors are pushing a case that the
police hierarchy thinks should go no further, my informants told me, they will find
that their oficios never arrive, witnesses are never home, documents cannot be
found, and police officers cannot be spared to carry out their requests. In São Paulo,
where ordinary investigations are carried out by the civil police, I did not hear similar
complaints.
In Salvador I found clearer evidence that the rule is enforced in both official and
unofficial instances, in ways both subtle and forceful. The occasional investigators
and prosecutors who are serious about investigating these cases are subject to career
sanctions. Marília Veloso, a former prosecutor in the military justice system, related
216 daniel m. brinks
in a personal interview in 2001 that her attempts to pursue violent military police
officers were impeded by her superiors, until she quit out of frustration. Nilton José
Costa Ferreiro, a former police investigator with the homicide division, also quit.
Costa Ferreiro was actively pursuing an investigation into military police participa-
tion in extermination groups. He had weathered various death threats and other
forms of intimidation by the police; he left when his efforts to investigate were met
with the warning that he himself would be prosecuted for abuse of power if he
continued (de Oliveira et al. 2000).
In summary, in Salvador da Bahia, there is evidence that a general rule permit-
ting the relatively indiscriminate use of force by the police is enforced. Prosecutors
and judges who seek to enforce the law are harassed not only by the police but also by
elected officials and their superiors in the legal system itself. The proof is less clear-
cut in the case of Buenos Aires, but there is some evidence that both the police and
the oversight institutions take measures to impede the effective prosecution of cases
in which the police have killed someone they perceive to be a violent criminal. In
the case of São Paulo, I have thus far presented little evidence that the pattern
detected is due to the application of an alternative rule of decision. Thus it is still
possible that in these cases, the failure to enforce the law is due to weak courts or
prosecutors who are unable to gather the evidence necessary for a full investigation
and prosecution (though it is difficult to explain why this weakness would affect so
peculiarly the Violent Victim cases). For further proof we will have to turn to the
remaining questions.
A closer look at individual cases suggests that the negative outcomes are not the
result of the legal system’s failure to observe the violations in question. The killing of
a person believed to have taken up arms against society often goes unpunished
whether or not all the information that might be needed to produce a conviction is
present.
In São Paulo, it is the case of Regiane Dos Santos that makes this point most
clearly. Regiane and her husband attempted to rob a home, taking the family hos-
tage when the police interrupted the robbery. In the subsequent exchange of gunfire
the police killed her husband and one of the hostages, a child. At this point, Regiane
surrendered herself and her gun to the owner of the house. The police entered the
house at the invitation of the owner, took Regiane to the bathroom, and summarily
executed her. The policemen were tried and acquitted on the basis of self-defense.
According to the attorney who represented Regiane’s children and mother, the
prosecuting police killings in brazil and argentina 217
judges blamed Regiane for the death of the hostage and ignored the testimony of the
owner of the house, who said she had surrendered and relinquished her gun before
the police took her away to be killed. The court of appeals affirmed the acquittal.∞∂
In Argentina there are several similar cases. The best known is probably the ‘‘Villa
Ramallo’’ case. In 1999, the police surrounded and incapacitated a car carrying four
people away from a bank robbery. They pulled the occupants out of the car and
executed all but the sole woman in the group. One of the three executed was a
robber, but the other two were hostages—employees of the bank. The same evidence
supports the conclusion that all three men were killed while unarmed and defense-
less after being pulled from the car. But the prosecution proceeds only as to the
killing of the two bankers, not the killing of the hostage-taker, who, while undoubt-
edly an unappealing individual, was also entitled to the benefit of due process.∞∑
The evidence from Salvador da Bahia is even more compelling, as the fate of
several well-publicized cases argues for an alternative rule of decision rather than a
failure of enforcement resources. One such case involves the killing of Robélio
Lima dos Santos. On October 11, 1999, Robélio was apprehended after he and three
others committed a bank robbery that resulted in one police officer being seriously
wounded. Robélio was photographed as he was handcuffed and placed in the back
of a police wagon. In the photograph, taken from no more than six feet away, one can
clearly see that he has only a single wound in the pelvic region. Some time later he
arrived at the emergency room of the local hospital. He was dead, shot in the chest
with at least two different weapons.∞∏
The local press published an editorial reporting the justification offered by other
police officers in defense of those who had killed this bandido. They noted that the
arresting officers were aware that one of their colleagues had been shot and seriously
wounded, and that it was likely they had acted out of vengeance. This attitude is
‘‘wrong, but accepted at an emotional level,’’ the officers argued. The editorial
concludes: ‘‘Policemen do not act as these Bahian ones did if they do not feel
protected.’’∞π The four police officers who were in the car when Robélio was mur-
dered were initially arrested for the crime, but when last I checked (nearly three and
a half years after the event), no convictions had resulted. So far, it does not seem that
the police officers involved in this case misjudged the normative expectations of
their superiors, or the legal system as a whole.∞∫
A more typical case, and one that received considerably less media attention, is
that of Sérgio Silva Santos, a physically handicapped youth who lived in one of the
favelas around Salvador. On January 22, 1999, five police officers on midnight rounds
decided to question a group of men standing around a street-side vending post. One
of the police officers had drawn his gun as he approached the men, and he acciden-
218 daniel m. brinks
tally fired a shot, wounding Sérgio in the neck. Not knowing what to do, they loaded
Sérgio, who was still alive, in the police car, took him to a remote region, debated
briefly about their course of action, and executed him, with at least two officers
taking part in the actual shooting. Then they placed a gun in Sérgio’s hand and
prepared a report that said he had died in an exchange of gunfire with the police.∞Ω
Eventually, confronted with witnesses to the initial wounding, the victim’s physi-
cal disability, and other damning evidence, one of the five police officers disclosed
what had actually happened, and testified to the entire sequence of events. In spite of
all this, at last report the accused are still employed by the police, and remain free.
The judge’s office reports that the difficulty in moving forward with the case lies in
locating the witnesses to the initial event. This should come as no surprise, as in April
2000 we read that ‘‘one of the main witnesses in the killing of the physically handi-
capped Sérgio Silva Santos . . . was beaten to death in Nordeste de Amaralina, the
neighborhood where he lived.’’≤≠ As noted below, the less publicized killings are,
ironically, surrounded by much less violence. There are few complaints, even fewer
investigations, and the cases are routinely ignored.
While no prosecutor or judge I interviewed stated directly that a police officer
should not be convicted under these circumstances, in many of these cases there is
no doubt that the violation of the formal rule was observed by the system, and still
there was no sanction, suggesting the application of an alternative rule of decision.
The decisions themselves are couched in more general language and use formal
legal categories to decide adversely, citing a lack of evidence to support a finding that
the police acted in excess of their legal authority. At worst, the courts simply fail to
rule at all. In Buenos Aires and Salvador da Bahia, there are cases that simply
languish until they are forgotten or until the statute of limitations expires. Moreover,
it is clear that in many of these cases, judicial officials have not undertaken the most
minimal efforts to uncover what really happened, despite gaping inconsistencies in
the police version of events. In short, the evidence suggests that the courts will not
convict in these cases even when the available evidence abundantly requires it.
As noted in the case of the Salvadoran bank robber, the police show considerable
reliance on, and expectation of, the lack of enforcement of the official rule in cases
involving violent criminals. In many cases they carefully create the appearance that
the victim was a menace to society. For more than half the cases in Argentina and
prosecuting police killings in brazil and argentina 219
nearly two-thirds of all cases in Brazil, I was given some indication—either because
judicial proceedings uncovered police deception or because advocates for the victim
made a concrete allegation—that the police had staged a confrontation after the
death of the victim, planted a gun, threatened witnesses, produced false forensic
reports, or tampered with the process in some similar way. The methods are diverse
and creative: one former policeman testified that the São Paulo police will place a
gun in the dead victim’s hand and pull the trigger so the skin will show gunpowder
residue; in one of the cases, the police shot their own car to simulate a confrontation;
in several, they were observed transporting the victim’s body to a different place
where they could stage a confrontation.
Often the evidence of tampering is irrefutable, as when a second report confirms
that the shots entered the back of the head, rather than the front of the body, or when
a gun attributed to the victim is traced to the comisario who had charge of the
operation. Other times it is simply the claim of an advocate for the victim. Some-
times the evidence is more subtle, but nevertheless persuasive, as in the case of
Darcy Ferreira dos Santos and Fábio Mário Saraiva Rodrigues. The indictment
against the São Paulo police officer who killed these men was dismissed, on the
grounds that they had entered a bar, drunkenly confronted the police officer, taken
out guns, and started shooting. The ability of the policeman to successfully repel the
aggression was somewhat surprising, since the two had allegedly started shooting
first. But the exculpatory evidence was provided by the owner of the bar, who
claimed to witness the event, and by the owner of an adjacent bar, who said the
victims had indeed been in his establishment drinking and brandishing guns imme-
diately before the incident.≤∞ That is typically as far as the investigation goes.
A closer look casts strong doubt on this version of events. The bar owner who
provided the exculpatory evidence is a relative of the police officer. The victims’
relatives, who talked to the police but were not called in by the prosecutor, claim
that the two young men left home only fifteen minutes before the shooting, were not
drunk, and were unarmed. The details of the autopsy report, which can be found in
the case file but are not discussed in any of the prosecutor’s filings before the judge,
show that in addition to one and two frontal wounds, respectively, each of the victims
had a gunshot wound in the same spot in the middle of the back, suggesting a coup
de grâce. The motive for the execution becomes clear when we learn that one of the
two men was on conditional release pending trial in the killing of a police officer. An
aggressive prosecutor could have uncovered these facts, ordered expert reports, and
put together at least a semblance of an effective prosecution. Instead, one wonders
why the prosecutor bothered to indict at all, when the indictment accepts without
220 daniel m. brinks
question and repeats the allegations that the victims initiated the shooting, and
makes no reference to any inconsistencies between the facts and the official story.
The outcome is predictable.
In Salvador da Bahia, the police also behave in a manner suggesting that they
know and rely on this rule: they are especially casual about killing, so long as it is
within the course of their duties, and seem to expect impunity. Witnesses and
relatives often report that police officers accused of a killing openly taunt them, to
the point that neighbors and relatives discourage complainants from going forward,
and the complainants themselves desist. The case of Sérgio Silva Santos, the dis-
abled youth who was first wounded and then summarily executed, makes this point
clearly: the police preferred to kill someone in cold blood, in an attempt to set up an
‘‘ordinary course of duties’’ killing, than to be caught having wounded someone by
mistake. Moreover, it was not one panicked police officer who did this, but no less
than five, after some deliberation. Similarly, it was not one enraged police officer
who shot and killed Robélio Lima dos Santos in the back of the police wagon, but at
least two, while two others were in the car with them. These are not individual
‘‘rogue cops’’; they are acting in accordance with accepted and established patterns
of conduct.
Thus far all the evidence suggests that the permissive rule is quite broad in
Salvador. But some evidence to the contrary suggests the permissive rule actually
applied by judicial officers is limited to killings committed in the course of official
police duties. The case of Heloísa Gomes dos Santos and her partner, Manuel
Ferreira dos Santos, shows that the police feel that violence is necessary to keep
information from the courts when the killing was committed in furtherance of
criminal activity on the part of the police. Manuel’s son Valdemir was killed by a
group of military police officers when he resisted attempts at extortion. Heloísa and
Manuel carried out a very public campaign to bring these policemen to justice.
They brought their complaints to the media, spoke to and joined various human
rights organizations, and spoke to state and federal legislators about police abuses in
Salvador. They even identified the alleged killers by name.≤≤
Soon after they began their campaign, Heloísa and Manuel began to receive
death threats. On June 21, 1998, they were shot to death as they sat in their car in front
of the hospital where Heloísa worked as a nurse. Since then, of the four witnesses
who testified against this group of policemen, along with Heloísa and Manuel, two
were murdered, one disappeared and is presumed murdered, and one is in hiding
under the auspices of Bahia’s witness protection program. There have been no
convictions in connection with the original extortion attempts, the murder of Man-
uel’s son, the murder of Heloísa and Manuel, or the murder of the other three
prosecuting police killings in brazil and argentina 221
witnesses. The use of such extreme measures to impede the prosecution is clear
evidence that the implicated police officers thought they could not rely on official
complicity but had to take self-help steps to ensure their impunity. In short, the
evidence suggests the application of a rule that is broad but not unbounded in
Salvador, and is limited to Violent Victims in São Paulo and Buenos Aires.
We might expect the informal rule that permits killings by police to be the hardest
to change in Salvador da Bahia, where it seems to be so widespread and supported by
so many at all levels of the state. But it is precisely here that there is some evidence
of a change in the rules, as evidenced by a few convictions in cases involving minors
over the last six or eight years. The Center for the Defense of the Child and Ado-
lescent (CEDECA), an organization fighting violence against children in Bahia,
selects particularly egregious cases involving poor children from marginal neigh-
borhoods and attempts to pressure the legal system to respond effectively. An inter-
view with its director suggests an instance in which the informal rules of the game
changed, in response to a change in the enforcement structure of the formal rules. At
the same time, a gross disparity of power between those who were negatively affected
by the new rules and those who benefited from them prevented a dramatic change in
actual outcomes, largely preserving the behavioral regularity.
Before the creation of a specialized juvenile justice system, police killings of
children and adolescents were open and frequent in Salvador. Many of the killings
fit the mold of the informal institution described above: a kind of social cleansing
carried out by, or with the cooperation of, the police against ‘‘undesirable’’ social
elements. In 1992, for example, a military police patrol confronted a band of youths
that engaged in petty crime in Liberdade, a peripheral neighborhood of Salvador.
The next day the police returned out of uniform and picked up two children they
suspected of belonging to the gang. One day later, the parents found their children’s
savagely mutilated bodies. The parents knew who had picked up the children, and
these police officers were initially charged with the crime. But the file on the case
mysteriously disappeared from the court records, and the case has gone nowhere in
ten years.≤≥
In another case, in 1991, a child from one of the invasões (shantytowns) around
Salvador tried to beg a ride from a bus driver. The driver refused to let him on, and as
he drove away the child threw a stone, hitting the bus on the side. The bus stopped,
and a policeman descended with his gun drawn and shot the child in the back as he
ran away. This case never even triggered a formal investigation by the prosecutor’s
222 daniel m. brinks
office and so did not make it into the courts at all. In 1992, another policeman
stabbed a child to death for hitting his dog with a stone. When representatives of
CEDECA confronted the police investigator in charge about his failure to open
a formal inquiry in this last case, his only response was: ‘‘When are people like
CEDECA going to establish a center for the defense of the police?’’≤∂
The stories are chilling; the killings were almost casual, perpetrators did very little
to hide their participation, the justice system did not investigate them, and no one
was ever punished for the crimes. CEDECA’s director told me in May 2001 that the
organization’s first systematic investigation revealed that eighty children were mur-
dered in 1991 in Salvador, most of them poor. Of those cases, only eleven led to a
formal judicial proceeding; only one of these led to a trial, and that single trial ended
in an acquittal. Though he could not say exactly how many, he noted that the police
were involved in a large percentage of these homicides. At this point the police made
little or no attempt to hide the facts, and the operative rule by all appearances
dictated that these crimes would not be prosecuted.
But the legal apparatus that attended these crimes in Salvador da Bahia was
reformed following the passage in 1990 of the (federal) Estatuto da Criança e do
Adolescente. This statute authorized the creation of the Vara Criminal Especiali-
zada da Infância e da Juventude, a juvenile justice system that acts in criminal cases
involving acts by or against minors. The investigators, prosecutors, and judges in this
system work only on cases involving minors.
According to the director of CEDECA, since cases began to be processed by the
new system, beginning around 1992, there has been a marked difference in the way
child homicides are treated. Whereas before it was difficult to find a police investiga-
tor, prosecutor, or judge who might show even a minimal interest, now the prosecu-
tors and judges, who deal only with children, are much more aggressive. Today, a
specialized investigative office (delegacia) opens formal inquiries for all the cases.
They are somewhat slow, they do not have a great many resources, and the inquéritos
often have errors, but at least the investigations are done in good faith, according to
CEDECA’s director. Similarly, one of the two judges in the juvenile system noted
many changes in the treatment of minors since the early 1990s, all of which permit
monitoring their safety from the time they come in contact with law enforcement.
Interestingly, the new official agency does two things. On the one hand, it offers a
more effective agency to control under-age crime, by ensuring there are facilities
that can process and house youthful offenders more quickly and effectively. This
neutralizes the proffered justification for dealing with these children extrajudicially
and undermines the claim that routinely permitting their murder is a necessary
substitutive institution. On the other, it strengthens the enforcement of laws protect-
prosecuting police killings in brazil and argentina 223
ing children from extrajudicial punishment by creating a corps of judicial and law
enforcement officers at all levels whose sole task is to ensure that the laws protecting
and governing children are respected. The performance of this corps will be judged
on the basis of how well it does this, and only this, task. Moreover, the statute itself
signals a national reaction to a perceived problem that would no longer be tolerated:
violence against children was singled out for special treatment and declared to be
unacceptable. Thus there is reason to hope that the informal institution permitting
the murder of socially marginal children might disappear altogether.
The behavioral regularity, however, may not disappear quite so quickly. As the
formal system became more aggressive in protecting the rights of children, a new
problem has come to the fore. Since the police can no longer count on benign
neglect by the justice system, the rate of threats and violence against complainants
and witnesses has vastly increased. Now, CEDECA’s director reports, relatives of the
victims come into his office in the initial days after the event, impelled by anger and
grief at the harm to or loss of a child, and asking the center to take some action to
ensure that the perpetrators are held accountable. But often they call back some days
later to withdraw the complaint, citing threats of violence against them if they persist:
‘‘His father has three other children to raise,’’ said a mother recently. ‘‘Even the judge
is afraid.’’ Witnesses cease cooperating with prosecutors and withdraw earlier state-
ments about police participation in the crime. In short, the police involved in these
killings now exploit the weaknesses of the system—primarily the vulnerability of the
lower classes that are the most frequent targets of violence—to close down the flow of
essential information to the system. As in the case of Heloísa Gomes dos Santos and
Manuel Ferreira dos Santos, violence against complainants increases when the cer-
tainty of impunity decreases.
The end result is a slight improvement in outcomes, but far less than what the
advocates of reform had hoped for. CEDECA can now point to several cases in the
last five or six years in which the courts convicted a police officer for killing a minor.
But there are still a large number of cases in which the prosecutions failed because
the police intimidated the witnesses or the complainants themselves. And CEDECA
can attest to an increase in the level of secondary violence, as the police respond to
stricter enforcement by taking more extreme measures to ensure that their illegal
behavior is not officially detected. This example shows why we should not extrapo-
late rules from mere regularities, or assume that a change of rules will necessarily
effect a change in behavior. It also sounds a cautionary note about the possible
negative consequences of changing the rules in the face of a continuing and serious
imbalance in resources between those who benefit from and those who are con-
strained by the new rules. The failure to protect the new rights-bearers ultimately
224 daniel m. brinks
exposes them to more violence, as the rules constrain the behavior of those who are
accustomed to impunity.
conclusion
short-circuits due process and shows no regard for the physical integrity of those who
come in contact with the police. In the final analysis, then, the outcomes of these
cases would be different if they were judged according to the formal rules, and
therefore we should classify this informal institution as a competing one.
As this example shows, informal institutions can substitute for some formal in-
stitutions while simultaneously competing with others, and it may not always be easy
to determine precisely what role they are playing. As should be clear by now, infor-
mal institutions often prove difficult to identify, describe, and categorize. This only
highlights the valuable contributions that remain to be made as we struggle to
understand the missing links between formal institutions and unexpected outcomes.
chapter eleven
todd a. eisenstadt
During the 1980s and 1990s, Mexico transformed from a hegemonic single-party
regime into a competitive democracy, culminating in Vicente Fox’s victory in the
2000 presidential election. Although the transition brought far-reaching change to
the country’s formal institutional architecture (Eisenstadt 2004), it was also marked
by the use of distinctive informal norms of interparty bargaining. During Mexico’s
protracted transition, dozens of postelectoral conflicts were settled through infor-
mally negotiated concertacesiones, or gentleman’s agreements. Invoked in nearly a
dozen national high-profile cases—and scores of lesser ones—during the 1990s, these
informal bargaining tables became the principal arena of negotiation between the
governing Party of the Institutional Revolution (PRI) and the opposition National
Action Party (PAN). The PAN usually lost in the PRI-state’s famously corrupt elec-
tions, but it was given something back—an interim mayorship or, at the very least, a
proportional-representation city council seat—as a reward for continued participa-
tion. The preferences of the electorate were put aside, as mayors- and governors-elect
resigned, under pressure from the PRI, to make way for negotiated ‘‘interim’’ execu-
tives from the losing PAN and ‘‘plural’’ municipal cabinets were shuffled in closed-
door negotiations to include an opposition co-mayor or city manager.
Concertacesiones operated at the margins of—and in lieu of—formal institutions
(such as electoral commissions and courts) that were designed to resolve electoral
disputes. Indeed, the demand for informal postelectoral bargaining tables was rooted
in the ineffectiveness of formal electoral and judicial institutions, as well as in PRI
and PAN leaders’ recognition that informal institutions offered them much more
228 todd a. eisenstadt
flexibility in tailoring electoral outcomes to the parties’ mutual needs. Once the PRI
and PAN learned of the discretion afforded them by the informal institution, without
any of the inflexible legalism or transparency required by formal institutions, concer-
tacesiones supplanted formal electoral institutions for a decade. The evolution of
concertacesiones thus offers an excellent opportunity to explore the rise and fall of
a substitutive informal institution (Helmke and Levitsky, introduction to this vol-
ume; see also Lauth 2000). At a time when electoral courts were still widely viewed
with suspicion, concertacesiones provided politicians with a temporary, but highly
flexible, means of adjudicating electoral disputes. The bargaining tables did not
give way to reformed formal electoral institutions until the end of the 1990s, as
increasing political competition fostered more calls for clean elections and far-
sighted opposition leaders pushed the one-party state to accept ever-more constrain-
ing electoral rules.
The term concertacesión (originating in a Mexican slang combination of the
words for ‘‘concertation’’ and ‘‘concession’’) has gained such cachet as a concept of
Mexican political culture as to merit inclusion in the country’s most important
electoral dictionary. The definition given is an ‘‘act by which, in cases where official
electoral results do not correspond to a reality sensed and witnessed by the electorate,
the official winner steps down and cedes to the candidate of the party which was
really thought to have won . . . This process offers a political out to severe electoral
conflicts’’ (Martínez Silva and Salcedo Aquino 2002, 99). This definition exaggerates
the role of voters in the process. Rather than responding to the electorate’s percep-
tions of who won, concertacesiones were a product of the PAN’s growing leverage vis-
à-vis the PRI at key postelectoral moments, which the PRI used to activate PRI-state
discretion in installing losing PAN candidates in local office (at the expense of local
PRI leaders). Voters and opposition activists played a role in making concertacesión
credible in two ways. First, they ensured a sufficiently close race that the PAN could
claim fraud on political grounds (even if not on legal grounds). Second, they pro-
vided ‘‘bread and circuses’’ for a visible local postelectoral mobilization, which,
while not of primary importance in negotiations, helped make concertacesión easier
for the public to swallow.
The chapter examines the rise and eventual collapse of concertacesiones. It shows
how an equilibrium was reached in which PRI and PAN leaders possessed incentives
to continue trying their luck at informal bargaining tables, rather than binding
themselves to formal institutions, and how this equilibrium unraveled as the incen-
tives of both actors changed. I argue that concertacesiones evolved out of preexisting
informal behavior patterns employed by the PRI to co-opt allies and repress threats.∞
mexico’s postelectoral concertacesiones 229
During the 1980s and 1990s, they became an integral part of the PRI’s effort to appear
more democratic without actually opening the electoral system. The PRI negotiated
away ‘‘consolation prize’’ spoils from its banquet of riches, allowing the party, in the
waning years of its seven decades of continuous rule, to maintain the PAN’s interest
in participating in elections. PAN participation gave the PRI-state—deeply embar-
rassed in 1976 when the PAN withdrew the only opposition presidential candidate
and the PRI still only managed to get 82 percent of the vote—a cover of legitimacy.
For the PAN, complicity with the PRI-state ensured access to local governance,
which allowed the party to establish a record of accountability otherwise unheard of
in Mexico. It also spurred the PAN’s legislative advocacy of electoral reforms, which
slowly leveled the electoral playing field and made way for the PAN’s strategic shift
to supporting only formal institutions when negotiation with the PRI grew less
essential.
Successful concertacesión required the confluence of six necessary factors: (1) a
PAN loss, but in a close enough race that the victory could credibly be ‘‘given’’ to it as
runner-up; (2) a PRI-state willingness to engage in postelectoral negotiations when
the PAN held a bargaining chip of national interest to the ruling party; (3) a willing-
ness of the local PRI to sacrifice electoral victory to make good on the national PRI-
state’s promises to the national leadership of the opposition party; (4) a disciplined
opposition with national leadership that could deliver on its promises and could
faithfully negotiate on behalf of local candidates; (5) local protests and mobilizations
of a magnitude sufficient to lead the electorate to believe that concertacesión was the
only solution to intractable local conflicts unresolvable by other means; and, per-
haps most importantly for the analysis of substitutive institutions, (6) the willingness
of participants to subvert formal institutions by relying on informal ones, and to
abide by agreements reached at informal bargaining tables, despite the adverse
effects on the credibility of formal institutions.
These conditions existed between the PRI and the PAN, but not the leftist Party of
the Democratic Revolution (PRD), from the late 1980s to mid-1990s. While the PRD
and other left-wing parties were involved in hundreds of postelectoral conflicts
during Mexico’s democratic transition,≤ the PRI, whose strategy was to co-opt the
conservative PAN and repress the left, utterly refused to engage with the PRD
national leadership. In any case, the national PRD lacked the discipline to deliver
the compliance of its local activists.≥ Hence, it was the PAN, which prioritized
winning regional governance positions in the short term as a segue to national
governance in the longer term, that benefited from concertacesión. PAN leaders
readily undermined formal institutions throughout the 1990s, but with the zealous
230 todd a. eisenstadt
belief that, by accepting what they called ‘‘partial restitution’’ of electoral injustices
now, they could pave the way for truly autonomous and transparent electoral institu-
tions (Eisenstadt 2004, 177–79).
Once the PAN had leveraged its postelectoral prizes to create a more competitive
party and to force the creation of more effective formal electoral and judicial institu-
tions, it abandoned postelectoral bargaining tables—even though the PRI desper-
ately wanted to keep them open. The demise of concertacesión can thus be traced to
two requisite factors: first, the growing credibility of formal institutions and, second,
the defection of the PAN, which reordered its preferences after winning the presi-
dency for the first time in 2000. Thus, when the PAN was strong enough to gain
national public office without the assistance of informal bargaining tables, and when
electoral commissions and courts gained credibility as genuine third-party enforcers,
the informal institution of concertacesión collapsed.
The chapter draws on original ethnographic research to elaborate the evolution
of concertacesiones. I break the rise and demise of concertacesión into four periods,
designated according to changes in the incentives of the national PAN and PRI
directorates. I conclude by situating the importance of concertacesión in the Mexi-
can transition, arguing that it was a crucial way station between the creation of
equitable formal electoral institutions and their acquisition of credibility (measured
as the observance of their dictates by relevant actors). Indeed, while concertacesiones
crowded out (and even delegitimized) formal institutions in the short term, in the
longer term they allowed actors to adjust their expectations regarding formal institu-
tions, moving from the perception that formal institutions were fraudulent and
inferior to ad hoc settlements, to a period of dueling focal points in which both
formal and informal institutions were used, to full compliance with formal institu-
tions and delegitimation of substitutive informal institutions. Hence, the PAN’s
strategic choice to use informal institutions in conjunction with constant efforts to
bolster the formal ones helped ensure that concertacesiones would eventually be
replaced by genuinely autonomous formal institutions.
The origins of concertacesión can be traced back to the heyday of the PRI’s
‘‘soft’’ authoritarian regime. The PRI-state’s legitimacy after its 1929 consolidation
stemmed from an ideological mission of institutionalizing objectives derived from
the demands of the Mexican Revolution: land reform, labor rights, and political
enfranchisement. By the 1970s, that platform had been diluted into symbolism and
pageantry, with few public policies attached. Nevertheless, the PRI maintained a
mexico’s postelectoral concertacesiones 231
corporatist system that mobilized citizens in official labor and peasant unions, and
managed a high level of popular support based on appeals to the ideals of a Revolu-
tion decades away. As long as the party-state could offer constituency services—albeit
discretionary and conditional—citizens would support (or at least acquiesce with)
the ‘‘catchall’’ PRI. Although the PRI-state at times violently repressed those, usually
on the left, who refused to go along with the system, the success of corporatism was
rooted in the effective co-optation of potential challengers. Thus, Mexico’s formal
electoral institutions, which were democratic on paper but corrupt and biased in
practice, coexisted with informal patterns of bargaining and co-optation.
Among the informal authoritarian institutions that emerged during this period
was the interim municipal council, a mechanism for resolving the many intraparty
conflicts that emerged after controversial candidate selection processes (Langston,
this volume). In municipal councils, interim mayors and/or city council members
were selected according to purely political criteria, rather than based on who won
the corrupt (but always held) popular elections—a pattern that would later be repli-
cated in concertacesiones. Hundreds of municipal councils were formed between the
1930s and the 1980s. By the early 1980s, they were governors’ legal solutions of choice
for postelectoral conflicts.∂ Between 1978 and 1981, for example, some twenty-five
plural municipal councils were formed after controversial local elections—a period
in which only 41 opposition victories were allowed nationwide (Assad and Ziccardi
1988, 41). By comparison, during the three-year period between 1990 and 1992, fifty-
seven municipal councils were formed, but 202 opposition victories were recorded,
and between 1993 and 1995, only nineteen municipal councils were created, com-
pared with 557 opposition local electoral victories.∑ During this period, postelectoral
conflicts were responsible for the creation of some 70 percent of the municipal
councils, while internal PRI conflicts provoked authorities to declare 10 percent and
other causes (e.g., corruption scandals, formation of new municipalities) triggered 20
percent.∏
While concertacesión was not really established until 1989, when the PAN ‘‘won’’
its first interim governorship in Baja California and its first mayoral substitution
in Mazatlán, two precursors to postelectoral conflict processes became routine by
the early 1980s. First, losing parties staged mass demonstrations, blocked freeways,
burned ballots, occupied public buildings, declared ‘‘parallel municipal govern-
ments’’ to sabotage public works by the winning party, and initiated mob brawls
when demonstrations were dissipated by local authorities or challenged by PRI
counterdemonstrations. Second, to ensure at least a modicum of electoral competi-
tion, the PRI-state began to concede some power, especially at the local level. Thus,
municipal reforms in 1977 and 1983 guaranteed proportional-representation city
232 todd a. eisenstadt
council seats for opposition parties and devolved to the localities a little more control
over the distribution of public resources. These reforms—together with the relegali-
zation of communist parties—generated unprecedented opposition interest in local
elections and gave rise to new patterns of participation and protest.
This surge in opposition political activity was fueled by economic crisis and
mismanagement, which shrank the PRI-state’s support base significantly during the
1980s and triggered important new challenges on both the right and the left. Al-
though the regime ignored the left’s postelectoral demonstrations, it paid attention to
those by moneyed PANistas in Mexico’s industrial north, who conveyed their grow-
ing opposition to the PRI’s statist economic model by sending tens of billions of
dollars out of the country (Barkin 1990, 109). The PAN, in existence since 1939, had
made strong electoral showings in isolated cases in the 1960s and 1970s and had
been stripped of several mayoral victories by the PRI’s legendary electoral fraud
machine—such as in the notorious Mexicali and Tijuana races of 1968, where the
PRI-state annulled elections rather than declare PAN victories, and a corrupt gover-
nor’s race in Yucatán in 1969 (Eisenstadt 2004, 167–72). In response to the fraud, the
PAN withdrew candidates from several local elections in the early 1970s, and the
party’s refusal to nominate a candidate to oppose the PRI in 1976 was the lynchpin of
defiance, which, along with the need to channel into the electoral arena insurgents
on the left (Klesner 1988, 391), catalyzed the regime’s granting of electoral reforms in
1977. But the reforms did little to change the tilted electoral playing field, and the
PRI continued to resort to fraud whenever even staged elections went awry. More-
over, the PRI’s domination (and corruption) of electoral bodies and the judiciary
meant that formal institutions lacked the credibility to act as third-party arbiter in
disputed elections. Consequently, the PAN, like the left, turned to protest. Although
the PRI would eventually domesticate these postelection mobilizational strategies
into concertacesión, the PAN launched a few genuine, widespread and disruptive
social movements in the mid-1980s, attracting cross-class support and international
media attention.
Meanwhile, imaginative opposition leaders, eager to appropriate claims of real
legality from the PRI-state’s arsenal of ‘‘window-dressing’’ formal judicial institu-
tions, established propaganda ‘‘people’s courts,’’ highlighting differences between
opposition positions and those of the PRI-state. By framing electoral justice as com-
monsense arbiters of intuitive ‘‘fair play,’’ in contrast to the formal institutions’ un-
intelligible and even suspicious legalisms, regime critics could claim that their
admittedly biased ‘‘courts’’ were no further removed from reality than those con-
structed by the official arbiters. The protests began in Puebla in 1982 with a poorly
publicized effort by the PAN to take election certification into its own hands. The
mexico’s postelectoral concertacesiones 233
party fared better in Chihuahua in 1986, where a ‘‘People’s Jury’’ of local notables
was formed by critical (pro-PAN but allegedly nonpartisan) opposition civic groups,
to monitor elections and issue the predictable conclusion that the election war-
ranted annulment due to its ‘‘great number of irregularities and vices’’ (Lau 1989,
87).π Another similar ‘‘People’s Electoral Court’’ was formed in Michoacán in 1989,
which declared, contrary to the state electoral commission, that the PRD (chartered
that year to unite the previously disparate left) had won a vast majority of the state
legislative races (Zamarripa 1989). Also along these lines were the ‘‘Electoral De-
fender of the People of Chiapas’’ and the ‘‘Electoral Tribunal of the People of
Chiapas,’’ which consisted of local and national notables convened through non-
governmental channels to ‘‘judge’’ the 1994 governor’s race (Vargas Manríquez and
Moreno Corzo 1994). Headed by an actress rather than a jurist, the publicity-seeking
intentions of this body were more transparent than its administration of justice. But
the opposition made its point.
These informal institutions of ‘‘cardboard commissions and courts’’ galvanized
public opinion, but held little sway with the authoritarian PRI-state until the left and
right united. For a fleeting moment in Chihuahua in 1986, and for a few resounding
weeks in 1988, the PRI-state faced the threat of a grand coalition against its decades of
ossifying ‘‘catchall’’ populism, but with dwindling state resources to distribute. After
the 1986 Chihuahua governor’s race was stripped from the PAN by the PRI-state, the
moralistic patron of Mexico’s recalcitrant left, Heberto Castillo, participated in PAN
marches and protests and argued that the PRI-state had a choice: ‘‘Either it respects
the popular will or else we engage in violence. I am ready to die’’ (L. Gómez
1991, 372). Castillo stepped aside as a 1988 presidential candidate to make way for
Cuauhtémoc Cárdenas, the left-of-center scion of one of Mexico’s most fondly
remembered generals and presidents. Although the PRI-state summoned all its elec-
toral ‘‘alchemy’’ (to use the common Mexican slang term) to push candidate Carlos
Salinas’s vote total to 50 percent (to Cárdenas’s 31 percent), Cárdenas was perceived
by many and, perhaps most importantly, by 1988 PAN candidate Manuel Clouthier
(who placed third, with 17 percent of the vote) to have been robbed of the presi-
dency. Clouthier temporarily bucked the conservative PAN establishment to join
Cárdenas in decrying the results (Loaeza 1999, 461).
The nightmare of a left-right coalition against Salinas’s fragile credibility—placed
in the service of Cárdenas’s opposition to the Salinas (and the PAN’s) neoliberal
platform—was said to have motivated many of the president’s actions, as well as those
of the conservative PAN hierarchy (Loaeza 1999, 462; Eisenstadt 2004, 175–82, 262).
The common threat posed to the PAN and PRI and the realignment of these parties’
preferences were propitious to the golden years of concertacesión. In siding with the
234 todd a. eisenstadt
PRI in an alliance for Mexico’s economic opening, after dalliances with the PRD
that only increased the PRI’s zeal to compromise and work with the PAN, the PAN
evolved from a more passive party seeking to win (or even just run in) local races,
publicize its conservative platform, and pursue only gradual and stable change, to
one threatened by the left and intent, first, on blocking radical change (by collaborat-
ing with the PRI-state in an anti-PRD alliance) and only secondarily on winning
local races. In period II (1988-96), starting immediately after the threat of the right-
left coalition first emerged, the PRI was marked both by the PAN’s aversion to radical
change and by its new partner’s desire to implement compatible economic reforms.
The 1988 election was so fraudulent that, even with vast majorities in both legisla-
tive chambers, the PRI-state needed PAN complicity in the Electoral College (the
lower chamber of Congress) to ratify President Carlos Salinas’s victory without
extreme embarrassment. Thus, during the summer-long postelectoral controversy,
Salinas dramatically increased the payoffs of concertacesión. The president-elect, in
need of PAN support against his PRD challenger, signed a letter of intent with the
PAN promising to enact the PANista-backed electoral reform in exchange for his
opponents’ complicity in certifying the 1988 election. While the PRD did not join
the PAN’s enduring electoral reform crusade—and then halfheartedly—until the
mid-1990s, the PAN seized its opportunity in 1989 not only to push its legislative
agenda but also to force Salinas to, in effect, rescind local PRI victories and throw
them to the opposition (Eisenstadt 2004, 176).
The PAN’s escalating costs of co-optation were also driven by the fact that the PRI
no longer possessed the two-thirds majority in both legislative chambers required
for constitutional reforms. In addition to needing a coalition partner and an oppo-
sition party legitimizer, Salinas also personally feared and deplored PRD leader
Cuauhtémoc Cárdenas. As late as 1994, Salinas was thought to have an almost
irrational fear of the PRD,∫ and that party’s early-1995 insistence that Salinas’s suc-
cessor, Ernesto Zedillo, resign so that the PRD might lead a ‘‘government of national
salvation’’ hardly endeared it to that president either.
The choice was thus clear to the authoritarian incumbents by the end of the
1980s. Either they would have to take their chances on further reforming the elec-
toral institution monster they had created, and at least gain a partner with whom to
finalize Mexico’s neoliberal economic reforms, or they would have to rely on their
increasingly belligerent regional PRI machines to defeat the antiregime PRD. The
PRI-state consciously chose the former—concertacesión. In 1989, Salinas sided with
mexico’s postelectoral concertacesiones 235
the PAN’s neoliberals over the populist machine bosses in his own party, and the PRI
leadership did not backtrack until two years before the 2000 electoral watershed,
when some activists realized that the party’s electoral firewall had been dismantled
and that its era of complete hegemony was over. Mexico’s most notorious concertace-
siones followed this 1989 move (table 11.1), as the substitutive informal institution was
used to supplant formal electoral courts and commissions on dozens of occasions,
including a half-dozen with profound national implications.
Some of the concertacesiones were shocking in their subversion of formal insti-
tutions—and the popular vote—to whitewash PAN-PRI deals. Several agreements
stand out during period II, but the most blatant case of subversion that I docu-
mented, and the one that set the standard for concertacesión as a substitutive institu-
tion, was the 1993 election in Mérida, Yucatán. The problem began when the local
chapter of the PRI in Mérida declared its candidate victorious, prompting the PAN
to launch an extensive postelectoral mobilization led by Diego Fernández de Ce-
vallos, the party’s just-named 1994 presidential candidate. In an effort to ‘‘undo’’ the
local PRI’s disobedience of the national PRI, which was increasingly more con-
cerned with garnering PAN cooperation than with who governed in Mérida, the
national Interior Secretariat composed an electoral court case to annul PRI votes so
that victory would pass to the PAN. After being flown to Mexico City for briefings on
the national PRI’s case, the Yucatán electoral court magistrates returned to Mérida,
heard the case, and did not revert the victory to the PAN (Eisenstadt 2004, 183–86).
Making no headway with the electoral court magistrates, the national PRI flew
the party’s Yucatán state legislators to Mexico City for discussions. According to
Rodríguez Lozano: ‘‘They spent a tormented twenty-four hours’’ being convinced to
overturn the mayoral race in the state electoral college. The resulting settlement, an
electoral college vote to designate the ‘‘runner-up’’ PANista to fill the term when the
PRI victor resigned under protest, clearly violated the Yucatán state constitution.Ω
The settlement was acknowledged as ‘‘illegal’’ and ‘‘antidemocratic’’ by the PRI
mayoral candidate who conceded, and by a national PRI negotiator who said he was
overruled by the Interior Secretariat.∞≠ It discredited the Yucatán state legislature
(which, when convened as electoral college, certified the election for the PRI but
awarded the post to the PAN) and the Yucatán state electoral court in favor of the
national and local directorates of the PAN and the PRI and the mayor of Mérida.
The official electoral college statement selecting the PAN runner-up, however,
lauded PRI candidate Orlando Paredes’s resignation as ‘‘an act of the highest civic
value, which puts the interests of Mérida and the State of Yucatán ahead of those of
any person, interest, or party.’’∞∞ Furthermore, ‘‘given this overwhelming circum-
stance, this Commission considers that to maintain political stability, social harmony
table 11.1
Deals behind Prominent Concertacesión Agreements, 1988–96 (Period II)
City or state election
(population Formal institution
governed, per 1990 said to have been
census, to nearest What the PRI What the PAN replaced by informal Probable mechanism
10,000) allegedly got allegedly got institutions of concertacesión
certifying federal elections was eliminated in all but the presidential race.∞≤ The
most important anti-concertacesión reform, however, was the 1996 incorporation of
local postelectoral disputes into the jurisdiction of the fortified federal electoral
court, subsumed under the judicial power. The Electoral Tribunal of the Judicial
Power of the Federation (TEPJF) owed its creation at least partly to the mayoral
concertacesión in Huejotzingo, Puebla, and to the aborted Tabasco gubernatorial
concertacesión (to the PRD this time) at the beginning of period III.
Concertacesión had never been the solution of choice for local PRI chapters,
which preferred to win elections at all costs, and then submit the results to sham
electoral commissions and politically driven electoral colleges, just as they always
had. During the Salinas administration (1988–94), the PRI’s local activists bided
their time, silently watching as the president and his national technocrats recruited
the PAN to help privatize and streamline Mexico’s previously bloated state, while
trading away the local elections they had fought so hard to win or steal.
In 1994, Salinas’s politically weaker successor, Ernesto Zedillo, sought to extend
the practice of concertacesión to the PRD in the aftermath of a fraud-ridden guber-
natorial election in Tabasco. As in the Salinas years, Zedillo’s interior secretary and
the national PRI engaged pragmatic PRD representative Andrés Manuel López
Obrador (Mexico City mayor and a leading 2006 presidential aspirant) in an effort to
end two months of PRD-led oilfield blockages, street protests, and work stoppages.
The Tabasco PRI had violated campaign spending limitations fifty-fold with impu-
nity, jammed the state’s electoral apparatus with biased officials who committed
flagrant election-day ‘‘engineering’’ (including a mysterious blackout of the vote-
tallying computer, which turned back on with 4 percent fewer votes cast), and
engaged in the usual ‘‘get out the vote’’’ promotions and petty ballot-stuffing (Eisen-
stadt 1999, 281). The PRD never expected electoral justice, but López Obrador,
taking his lesson from the PAN, explained: ‘‘Whatever we do, it will be construed as
acting outside the law, so we must proceed through strict legal channels [in addition
to extralegal channels] to avoid these criticisms. We must follow this legal course,
even as we mobilize citizens, knowing that it [the law] does not work.’’∞≥
This PAN-like dual strategy of legal challenge and mobilization (a departure for
the PRD, which had dedicated most of its effort to the latter) was particularly salient
because hundreds of Tabasco PRD activists marched to Mexico City and occupied
the city’s central square, threatening to interrupt Zedillo’s presidential inauguration
mexico’s postelectoral concertacesiones 239
(López Obrador 1996, 170). As dialogue over Tabasco lagged in January 1995, the
PRD forced the PRI to agree to at least ‘‘reconsider’’ 1994 postelectoral controversies
in several states in exchange for PRD participation in reform talks.
The Tabasco concertacesión ultimately fell apart, however. When word of a pos-
sible concertacesión reached Tabasco, local PRI leaders preempted the PRD by
taking to the streets themselves. The Tabasco legislature’s PRI leadership declared
that the local party had been abandoned by Mexico City, and the PRIistas agreed
among themselves to refuse any resignation tendered by the governor-elect and,
furthermore, to adopt the PRD’s mobilization tactics to get Mexico City’s attention.
They succeeded. When the ‘‘rebellion of the PRI’’ threatened widespread violence
and ungovernability, Zedillo rescinded the forced resignation of governor-elect
Roberto Madrazo (who had already been offered a cushy federal cabinet position in
exchange for his projected personal sacrifice) (Eisenstadt 1999, 285–87). The politi-
cally inexperienced Zedillo, apparently lacking the authority to impose such legal
shenanigans on the local PRI machine, distanced himself from the PRD’s concer-
tacesión gone sour and reaffirmed his staunch advocacy of the rule of law.
The Tabasco episode marked the emergence of a more moderate and pragmatic
PRD that was willing to use both formal and informal institutions. But more impor-
tant to the future of concertacesión, the ‘‘rebellion of the PRI’’ drew attention to the
growing breach between the incentives of election-driven local PRI activists and
their policy-driven national leaders. It also demonstrated to local PRI factions that,
unlike in the Salinas era of strong party discipline, rejection of national PRI dictates
was an option under Zedillo. Condition 4 of the six necessary conditions for concer-
tacesión outlined earlier—a disciplined opposition party with a national leadership
that could deliver on its end of informal bargains—had been violated, but by the PRI
rather than the PAN.
PAN activists, however, continued to view concertacesión as a viable, if second-
best, option during the mid-1990s. Taking advantage of Zedillo’s interest in including
the PAN in a definitive electoral reform, the national PAN activists found cause in a
small town where electoral court rulings were an abomination, even to reformers in
the PRI. With the rallying cry ‘‘democracy passes through Huejotzingo!’’ the na-
tional PAN shone light on sham formal institutional rulings not witnessed since the
1993 Mérida mayoral ruling by the Yucatán state legislature. Local PRI efforts to fix
the PAN’s 1995 Huejotzingo victory resulted in an electoral tribunal’s whitewashing
of the outcome by overturning the PAN, on shaky legal grounds, in favor of the PRI
(Eisenstadt 2004, 190–91). The national PAN mobilized, finessing the PRI mayor’s
resignation in favor of a PAN-selected interim mayor. The interior subsecretary,
Arturo Nuñez, acknowledged that who was mayor in Huejotzingo was deemed less
240 todd a. eisenstadt
tables turned, with the PAN-as-government getting to make the discretionary Solo-
monic choices—would undermine a decade of PRI-state rhetoric about the rule of
law and electoral transparency.
Farsighted PRI strategists, however, began to see, even before the 2000 electoral
defeat, that they were losing the discretion and control that concertacesión had given
them. During period IV, the PRI would learn that the PAN was no longer willing to
fulfill necessary condition 6 of concertacesión: unlike the PRI-state, the PAN-as-
government would not sully formal institutions with the credibility loss from concer-
tacesión. To the contrary, the new PAN administration was composed of concertace-
sión veterans—such as President Vicente Fox—who disliked negotiating votes at
bargaining tables almost as much as they disliked losing to PRI-committed fraud.∞π
Dislodged from the state apparatus that had nourished it for seventy-one years,
the PRI grew even more desperate in the months following its July 2000 presidential
defeat. Casting aside the formal electoral institutions it had tolerated—and even
promoted—as part of the PRI-state, the PRI’s national leadership backpedaled strate-
gically to the heyday of concertacesión with the PAN. Reminiscent of the PAN’s
threats to boycott Zedillo’s inauguration if its Monterrey mayoral candidate was not
recognized (another notorious concertacesión; see table 11.1), the PRI threatened to
boycott Fox’s inaugural if the PAN did not agree to a ballot-by-ballot recount of the
Jalisco election for governor, which the PRI had lost. Although initially the PRI and
PAN sat down at a bargaining table with the outgoing interior secretary to hammer
out a deal, the PANistas, entering office in a matter of days, stalled negotiations.
Having finally gained the upper hand, they referred the angry national PRI leader-
ship to the federal electoral court, where the PRI clamored for ‘‘relief ’’ to the elec-
toral magistrates, who, freed of any ties to their nominators (Zedillo and his Supreme
Court) were more interested in continuing to develop their growing reputation for
impartial electoral justice than in placating the PRI.
In sharp contrast to the Monterrey mayoral election of 1994, the federal electoral
court refused to whitewash the Jalisco gubernatorial election. After months of verbal
recriminations, the PRI’s poorly argued case (in which the party violated the appeals
process by changing its argument between the time of the initial hearings in the
Jalisco electoral court and the federal appeal) was ruled as mostly unfounded in
February 2001, leading to further—unfounded—attacks against the electoral court
242 todd a. eisenstadt
(Granados Chapa 2001). By the admission of their own legal advisor,∞∫ the PRIistas
were the last to realize that the electoral courts had become real, independent
arbiters, whatever the PRI-state’s intentions when it allowed for their creation.
The PAN had long used the idealized norm of electoral justice and parliamentary
persistence to push for creation of an electoral court.∞Ω By the late 1990s, the PRD
had also finally resolved to use the courts effectively as a check against the most
heinous regional PRI machine bosses, spearheading legal—rather than political—
efforts in 2000 and 2001 to overthrow PRI-rigged elections in Tabasco and Yucatán.
The federal electoral court’s inscrutably legal and politically dispassionate Jalisco
ruling confirmed its professionalism,≤≠ and exposed the widening breach between
the diehard local PRIistas’ efforts to cling to the past and the national policymaker
moderates’ escalating embarrassment with their own party. Rather than being pulled
together by an emerging leader, the internal division only pushed PRI partisans
further apart, prompting postponement of the election of a new national leadership
(originally slated for 2000) to 2002.
Also in 2000, the federal electoral court annulled that year’s PRI gubernatorial
victory in Tabasco. The electoral court cited several ‘‘indicators of electoral fraud’’:
electoral authorities’ insistence on opening all the preliminary results statewide
without justification, the dramatic PRI monopoly on television publicity and news
coverage, the location of contraband electoral materials such as ballots at a PRI-
contracted survey research firm, and compelling but circumstantial evidence that
Governor Roberto Madrazo had mobilized public resources for the campaign. As
the electoral court magistrates argued that they had been gradually amassing tools of
judicial activism to render the powerful verdict,≤∞ the political parties—and espe-
cially the PRI, accustomed to a servile electoral court—were shocked.
Having indirectly defied the PRI’s perhaps most powerful governor (the PRI’s
national president and a leading 2006 presidential aspirant), the federal electoral
court then directly challenged another PRI stalwart boss, Víctor Cervera Pacheco,
the governor of Yucatán. If the Tabasco ruling helped quell last-ditch efforts at
concertacesión, the court’s Yucatán actions in defense of formal institutions ended
most talk of concertacesión once and for all, especially since the Fox administration
wanted nothing to do with negotiating electoral outcomes or even any appearance
of intervention in local politics.≤≤ Governor Cervera Pacheco’s argument that the
electoral court’s intervention in December 2000 to select a slate of state electoral-
commission citizen counselors was a violation of federal law was unsubstantiated.
Contrary to widely circulated claims within Yucatán that the federal electoral court
overstepped its bounds by selecting Yucatán’s citizen counselors in Mexico City, the
federal electoral court was empowered to do so under Mexico’s Constitution. Judi-
mexico’s postelectoral concertacesiones 243
cial authorities insisted that unchallenged precedents existed for direct federal inter-
pretation of state electoral laws and of election-related state legislature acts.≤≥ The
original violation of the Yucatán electoral code was self-evident: the electoral com-
mission ombudsmen required approval by a four-fifths majority of the state legisla-
ture (or twenty of twenty-five members), and only fifteen legislators approved the
initial slate. In October 2000, the TEPJF ruled that the Yucatán legislature had
selected its electoral commissioners illegally and called for a new selection.
The Yucatán legislature agreed initially to uphold the electoral court decision,
reselecting the same slate of electoral commissioners, but with the needed twenty
votes. However, irregularities in this process prompted the electoral court to inter-
vene again to request stricter adherence to the original verdict, which the Yucatán
legislators refused. Cervera Pacheco and his legislative leaders were not deterred
by the contradiction of first accepting the electoral court decision and then back-
tracking and declaring it unconstitutional. Nor was PRI national president Dulce
María Sauri, the former Yucatán governor who resigned over the 1993 concertacesión
of Mérida with the PAN and who sided in 2001 with Cervera Pacheco against
Mexico City.
Waving the banner of state autonomy and even threatening secession from the
union, Cervera Pacheco found further political resonance in defying a late 2000
federal electoral court verdict to block the Yucatán governor from stacking the state
electoral commission in favor of his party in the run-up to the 2001 gubernatorial
election. His ‘‘yucatecos versus the Mexico City bureaucrats’’ campaign, resound-
ingly endorsed by 95 of the state’s 106 mayors, boosted the governor’s popularity and
caught the federal electoral authorities between two poor options: ordering en-
forcement of their verdict via public force, or failing to intervene, losing credibility,
and opening future decisions to second-guessing by emboldened caciques, or local
bosses. Fox, himself a victim of the 1991 concertacesión in Guanajuato (table 11.1),
insisted that the Yucatán legislature was responsible for following the federal elec-
toral court’s dictates and that his government planned not to intervene, since ‘‘the
yucatecos have the ability to resolve their own issues and I’m sure they are going to do
so’’ (quoted in Olvera Aguirre 2001, 5).
The proposed solution, which allowed federal authorities to bide their time while
Cervera Pacheco played out all options short of violence, was for the federal elec-
toral court to name a ‘‘legal’’ Yucatán electoral commission to compete with the
‘‘illegal’’ Yucatán legislature–ratified electoral commission. Not unlike the Puebla
1983 ‘‘People’s Tribunal’’ to publicize electoral fraud, the Chihuahua 1986 ‘‘Peoples’
Jury,’’ and the 1994 ‘‘Electoral Defender of the People of Chiapas,’’ dueling institu-
tional focal points were found to be the best means of distending conflict. However,
244 todd a. eisenstadt
the key difference between the Yucatán 2001 electoral commissions and the Puebla,
Chihuahua, and Chiapas ‘‘peoples’ electoral courts’’ was that this time the legal
authorities had opted for the duality as a means of stemming conflict, whereas, in the
prior cases, opposition parties and other societal interests had initially chosen to vent
conflict only through the parallel institutions. The resort to dual institutions pro-
vided a procedural symmetry.
The solution to the Yucatán conflict was political, as the legislatively appointed
electoral commission received the $4 million budget for conducting the May 2001
elections, but a second, federal electoral court–appointed commission received the
federal voter list (usually ‘‘lent’’ by the federal electoral authorities to each state’s
electoral commission). Still, Cervera Pacheco did not desist until the Mexican
Supreme Court ruled in April 2001 that combining the two electoral commissions
into one ‘‘super-commission’’ was illegal. The Cervera Pacheco–backed commis-
sion (formally named by the PRI-dominated legislature) finally stood down, to allow
the naming of a compromise electoral commission to organize elections some six
weeks hence.
The Yucatán conflict broke with previous concertacesión patterns in several ways.
First, the protesters who occupied the state electoral commission for two months
were all PRIistas, breaking the opposition-led protest dynamic of the past. Second,
the hard-liner–led national PRI was united with the local Cervera Pacheco machine
in opposing the federal electoral court ruling, against the PAN-led federal govern-
ment and PRI moderates. Third, contrary to past patterns of Interior Secretariat
intervention, the national government was loath to take sides. Fourth, the protesters’
ire targeted a legal decision, justified by an emerging field of law and legal institu-
tions, rather than the usual arbitrary concertacesión (or lack thereof, in PRD cases).
In managing to move the public discourse from debate over the arbitrariness of a
concertacesión to debate over the legality of an electoral court decision, federal
electoral authorities won even before the case was decided.
The electoral court’s decision to select its own Yucatán electoral commission
was all the more powerful coming on the heels of the landmark Tabasco guber-
natorial case. National PRI president Sauri inconsistently accepted the federal elec-
toral court’s annulment of the Tabasco governor’s race, though refusing to allow
federal electoral court meddling in her home state’s gubernatorial race, and the
national PRI largely heeded her position. A few moderates bucked the PRI leader-
ship, but most remaining PRIistas unquestioningly toed the party line. The contra-
dictions with decades of PRI rhetoric praising law, reason, and restraint by those who
had stewarded the development of autonomous electoral institutions (even while
seeking not to actually use them) were exposed. At their worst, the PRI’s new leaders
mexico’s postelectoral concertacesiones 245
appealed to the base instincts of mob rule; at their best, they enlisted the popu-
list anti–Mexico City themes of regionalism, federalism, and sovereignty to jus-
tify disobeying federal mandates. Either way, the powerful and unobstructed de-
fiance by the electoral court of what just a few years before had been ‘‘untouchable’’
forces signaled the final demise of concertacesión. Once in power, Fox and the PAN
were unwilling to indulge the PRI’s engagement with informal institutions, and the
success of Mexico’s electoral institutions meant that subversion via concertacesión
was destructive of the formal institutional architecture the PAN-as-opposition had
worked so hard to construct.
Leaders of the PAN made a moral choice during the peak of concertacesión
during the late 1980s and early 1990s. On deciding to seat the party’s members in the
Chamber of Deputies to certify PRI candidate Carlos Salinas’s irregular election as
president in 1988, the PAN party president, Luis H. Alvarez, set a precedent for
partial co-optation with the PRI, declaring that ‘‘we have to be capable of negotiating
and accepting partial and provisional formulas of transition’’ (Loaeza 1999, 462).
This was a controversial conclusion, as the early PAN activists—especially the party’s
distinguished lawyers and parliamentarians—were directly responsible for proposing
and stewarding the electoral reforms that the PRI-state had allowed. Whereas PAN
purists sought to fortify the credibility of formal electoral institutions only, Alvarez
and the pragmatists—who led the party from its early decades as a doctrinaire debat-
ing society to its politically crucial role as the engine of Mexican democratization—
settled for a second-best strategy. They would seek to create and fortify strong formal
electoral institutions in the long run, but accept informal bargaining tables in the
short run.
The PAN’s strategy, known as ‘‘partial restitution’’ of electoral improprieties
through bargaining tables, ran counter to the legal purists’ argument that jus ex
injuria non oritur (‘‘rights do not arise from wrongs’’). Clearly, participation in infor-
mal electoral institutions—especially substitutive ones, which undermined the cred-
ibility of formal institutions—was inconsistent with the PAN’s advocacy for the rule
of law. But by the early 1990s, the pragmatists had defeated the purists. Acknowledg-
ing the moral compromise involved, an interim 1991 PAN Guanajuato state governor
averred that concertacesión had been used to ‘‘partially restore’’ fairness after fraud
in that state had cost now-President Vicente Fox the governorship. According to a
PAN National Council release: ‘‘National Action considered the extreme dichotomy
246 todd a. eisenstadt
PAN quickly cast these institutions aside. The vital legacy of these substitutive in-
stitutions was in tipping the balance of forces toward the PAN even before the 2000
alternation, not in the design of their formal institutional counterparts. The PANistas
were too disciplined to allow the substitutive institutions to ever replace formal
institutions, which they always used in tandem with their substitutive analogues.
After a few iterations, expectations converged with those of PRIistas about how the
concertacesión bargaining tables would be run. But, apparently contrary to the PRI
(which sought concertacesión unsuccessfully several times in 2000 and 2001), the
PAN never considered concertacesiones as more than a second-best option and al-
ways kept the goal of perfecting the formal institutions of the rule of law within
its sights.
Of the six necessary conditions for concertacesión listed earlier, the final condi-
tion—that actors be willing to accept the subversion of formal institutions inherent
in their being substituted by concertacesión—thus provided the greatest impetus for
change. The parties’ incentives changed during each period in the evolution of
concertacesión, but the PAN did not lose sight of its primary goal of long-term stakes
in the creation of autonomous formal institutions, and maintained great discipline
in ‘‘settling’’ for partial restitution ad hoc while longing for systematic electoral
justice.
A broader theoretical point is worth noting. The substitutive concertacesión bar-
gaining tables did temporarily crowd out or undermine formal institutional resolu-
tions, as suggested by Helmke and Levitsky in the introduction to this volume, but
they did so only as long as the formal institutions failed to impart justice. By clearly
perceiving when they no longer needed concertacesión, the PANistas were able to
disregard the practice as a rung in the ladder they had already climbed. The PRI-
state allowed the PAN (and, after the mid-1990s, the PRD) reformers to gradually
empower formal institutions, electoral reform by electoral reform, in a feedback
loop that was also driven by the PAN’s improving electoral competitiveness, espe-
cially in traditional opposition strongholds (the same areas that had been tainted by
concertacesión).
By the time PRI strategists realized, in the late 1990s, that they had empowered
the electoral institutions beyond their ability to control them, it was too late. The
PRIistas did not realize (and probably could not fathom) that the PAN would stop
submitting to concertacesión and switch to a policy of exclusive reliance on formal
institutions at the earliest opportunity. The mutually reinforcing trends of improved
PAN electoral performance and fortified formal electoral institutional autonomy
allowed Mexico’s oldest opposition party and the electoral institutions it had imag-
248 todd a. eisenstadt
ined to prosper in tandem. When the formal electoral institutions showed them-
selves to be sufficiently autonomous and transparent to be worthy of compliance, the
post-2000 PAN-government vested itself completely in them, exposing the hard-
liners in the PRI as the only remaining proponents of concertacesión. To the PAN,
concertacesión was just a means to the end of powerful formal institutions, but to the
PRI’s hard-liners, it brought the end of their means.
chapter twelve
Informal institutions often emerge in the vacuum created by weak formal institu-
tions. In Latin America, formal institutions may be weakest with respect to the rule of
law. For centuries, segments of Latin American geography and society have existed
outside the reach of the state. As Guillermo O’Donnell (1993) has observed, large
swaths of Latin America may be considered ‘‘brown areas,’’ in which most citizens
are not protected by the courts and police. In the absence of judicial control, corrup-
tion and impunity are endemic, and conflicts are resolved and order maintained
through extrajudicial means. Such problems have captured the attention not only of
scholars. Policymakers are likewise concerned that the weakness of judicial institu-
tions severely inhibits the achievement of economic, political, and social progress.
The World Bank recently declared that justice administration in Latin America is
‘‘inefficient and ineffectual,’’ arbitrary, inaccessible to many citizens, underfunded
and lacking in modern facilities and practices, and not subject to external review. As
a result, public confidence in judicial systems is low throughout the region (Malik
and Martínez 1999, 827). The quality of Latin American justice administration is so
poor that even powerful economic interests are seeking alternatives in private arbitra-
tion fora. In this context, it is tempting to view informal justice institutions as uni-
formly benign—albeit imperfect—solutions to the demand for justice.
Yet, while informal justice institutions often emerge or persist in the vacuum
created by the weakness of state law, I argue that the relationship between informal
and formal legal institutions is more varied and complex than the conventional
wisdom suggests. Indeed, informal justice systems may fall into all four of the catego-
250 donna lee van cott
ries discussed in Helmke and Levitsky’s introduction to this volume. Although many
are fully consistent with state law or substitute for absent or incompetent state
institutions, others directly compete with penal laws.
Modifying slightly the typology developed in the introduction, I explore informal
justice institutions along two dimensions. The first is the degree of contact with and
coordination between a particular informal institution and the state. At one end of
the spectrum, informal authorities act as purely substitutive institutions, in that they
are wholly responsible for justice administration and dispute resolution. At the other
end of the spectrum, informal institutions complement formal institutions by sharing
responsibilities with the police and courts, particularly for issues not addressed by the
state. As the chapter demonstrates, most cases fall somewhere in between these two
ideal types. The second dimension is the propensity of the informal and state author-
ities to alter their institutions to adjust to the presence of the other. On this dimen-
sion, institutions may range from competing to accommodating. Where informal
practices violate penal law and no mutual adjustment occurs, the relationship may
be described as competing. However, instances of pure competition are decreasing
because more than a dozen Latin American states now legally recognize some
informal justice systems. In addition, community authorities are reshaping their
own norms and procedures to adjust to the new regime of ‘‘legal pluralism’’: the co-
existence of distinct legal systems in one geographic space (von Benda-Beckmann
1997, 1). In some cases, formal justice institutions are being accommodated to com-
munity cultures and customs, and the police, courts, and community authorities
work together to detain suspects, investigate offenses, and select the appropriate
venue for the administration of justice.∞
The two continua described above are neither mutually exclusive nor interdepen-
dent. A given informal institution can be classified on both dimensions (i.e., the
extent of the role of formal institutions and the degree of mutual accommodation
between formal and informal authorities), and the location of a particular commu-
nity’s informal justice administration system on one continuum does not necessarily
determine its location on the other.≤
I begin by examining diverse examples of informal systems of community justice
administration, explaining how and why these informal legal institutions emerged,
how they have evolved over time, and their relationship(s) to the state. I then ex-
amine efforts in the last decade to legally recognize informal justice institutions
and to link their jurisdictions to that of the state, and the reasons for the grow-
ing trend toward formal recognition. I also discuss the theoretical and practical
challenges generated by efforts to establish legal pluralism. Because the Andean
countries (particularly Bolivia, Colombia, Ecuador, and Peru, and Venezuela) are
the informal rule of law in latin america 251
most advanced in recognizing the authority of informal legal systems, they receive
more attention.
The norms and procedures of indigenous law vary so greatly among cultures,
even within regions of particular countries, that cogent generalizations about them
are difficult. Indigenous law encompasses ‘‘the uncodified concepts, beliefs and
norms which, within a given community, define prejudicial actions or crimes; the
selection of authorities and processes by which these should be resolved; and the
sanctions or resolutions decided and applied’’ (Sieder 1998, 98). For example, indige-
nous law typically determines who will resolve disputes within the community and
the procedures for hearing and resolving these disputes; the acceptable standards for
the use of coercion in domestic relations; and who should have access to agricultural
and pasture land within a community’s collectively owned territory.
Nonspecialists often receive an essentialized and static version of indigenous law,
which is purported to protect centuries-old traditions—a vision often propagated by
indigenous leaders themselves as a strategy to defend their autonomy. This view of
indigenous law also was propagated during the twentieth century by positivist state
officials, who view indigenous justice systems to be ‘‘customary law,’’ an essentially
static, subordinate, and less authoritative form of law. In fact, however, contempo-
rary indigenous systems are flexible and dynamic, and many practices and norms are
of recent vintage. Indigenous legal systems are constantly adapting to changing
circumstances (Orellana Halkyer 1998, 232; Sánchez Botero 2000, 224; Yrigoyen
Fajardo 2000, 198). A second misconception about indigenous law is derived from
the statements of some anthropologists and indigenous movement leaders, who
portray indigenous communities as harmonious, homogeneous, and unified collec-
tivities. In fact, indigenous communities typically are rent by internal conflicts, and
we must resist the temptation to romanticize them (Jackson 2002, 120).
252 donna lee van cott
Origins. Before the arrival of Europeans, the peoples indigenous to the Americas
established systems of self-government, including norms of acceptable behavior and
procedures for enforcing them. Thus, indigenous law has a special character with
respect to state law: its existence and authority precede the creation of the state (M.
Gómez 2000, 5). The imposition of colonial rule, however, had an enormous impact
on these systems. In some cases they were destroyed. Others were transformed and
adapted to colonial power relations. Colonial administrators tolerated the normative,
administrative, and jurisdictional activities of indigenous authorities for the manage-
ment of minor, internal matters that did not impinge on state or divine law (Yrigoyen
Fajardo 2000, 206). Indigenous authorities served as useful intermediaries between
colonial authorities and the native population and as an efficient means of indirect
social control. Some indigenous legal systems originated in the colonial period in
order to link community and colonial authorities. In cases where they were designed
and imposed by colonial authorities, over time, Indians appropriated and adapted
them, and made them their own.
Evolution. After independence, with few exceptions, such as Colombia and Gua-
temala, states influenced by liberalism and positivism outlawed indigenous legal
systems and enacted policies to forcibly assimilate indigenous peoples and destroy
legal distinctions among ethnic groups. Despite the formal ban, however, indige-
nous legal systems continued to operate where indigenous communities survived. As
a rule, the more remote the geographic location of indigenous cultures, the greater
the relative isolation in which such authority systems developed and the greater their
autonomy. In locations closer to state power structures, indigenous peoples incorpo-
rated more aspects of Western law and were more likely to coordinate their justice
systems with those of the state, particularly in cases involving outsiders or serious
offenses.
Indigenous legal systems survived for two reasons. First, they helped to perpetuate
and defend the autonomy of indigenous cultures and the sovereignty of indigenous
nations and their authorities. This function is distinct from some of the other infor-
mal institutions described in this book, which were developed by elite members of
the dominant culture. Second, state legal systems, in addition to discriminating
against the indigenous, tend to be inefficient, inaccessible, and culturally inappro-
priate for dispensing justice in indigenous communities. Indigenous justice institu-
tions emphasize resolving conflict and restoring social harmony, thereby strength-
ening the community’s system of reciprocal assistance and mutual responsibility.
Because they are based on commonly held indigenous values and norms, they have
greater legitimacy. They use indigenous languages and are socially, culturally, and
the informal rule of law in latin america 253
geographically closer to the communities they serve. They also act faster and are less
costly than the state system, since they do not require lawyers and there are usually
no fees (Ochoa García n.d., 13; Yrigoyen Fajardo 1999, 44–45).
The survival of indigenous law, however, is due as much to its adaptation to new
circumstances as to its retention of long-standing practices. Indigenous legal systems
underwent significant transformations in the twentieth century as they adjusted to
changing relations with outside actors, as well as changes within their communities.
Beginning in the 1960s, in many highland areas indigenous authority systems were
replaced with or subordinated to campesino unions. In the past decade or so, how-
ever, as indigenous identity has increased in prestige and become the subject of
advantageous rights, many indigenous authority systems have been revived (Yrigo-
yen Fajardo 2000, 204). The pace of change increased in response to the active
recuperation of indigenous authority structures, which was provoked by the emer-
gence of strong regional and national indigenous peoples’ social movements in most
Latin American countries in the mid-1970s and 1980s.
For example, in two Ecuadorian Quichua communities during the past decade,
communal assemblies have reformulated their internal rules, which now are re-
viewed annually (García Serrano 2000, 15). In one Quichua community, García
found that community authorities, together with representatives from the Catholic,
evangelical, and Mormon churches, had established a new justice institution—
the Autonomous Social Justice Commission—that was neither a part of traditional
Quichua culture nor contemplated in the 1937 Ley de Comunas. In fact, in the
Quichua communities of Ecuador that he studied in the late 1990s, García found
that the revived systems of indigenous justice administration had completely re-
placed the formal state system: lawyers and judges in cities near indigenous com-
munities that had recently strengthened their own systems reported a ‘‘total absence
of claims’’ by indigenous people in their offices and courts (García Serrano 2000,
15). Moreover, the superiority of indigenous justice systems to state fora was not
just based on their cultural appropriateness. García’s informants reported that it is
common for mestizos and Indians from communities lacking such systems to take
their disputes to the newly strengthened indigenous tribunals, particularly those dis-
putes that the state is unable or unwilling to resolve. As the president of the Unión
de Pueblos Chibuleos observed: ‘‘Even when the problem is only among mestizos,
they come here to resolve the problem and they leave tranquilly’’ (García Serrano
2000, 20).
Relationship to the State. Though not officially recognized by most Latin Ameri-
can states until the 1990s, indigenous law has long been influenced by the state. As
254 donna lee van cott
Orellana Halkyer (1998) shows in the case of Bolivia, many indigenous communities
are eager to learn more about state law and to obtain documents concerning legisla-
tion that affects them. Access to these laws often prompts changes that replace
traditional norms. Orellana Halkyer (1998, 233) also observes an increasing tendency
for indigenous authorities to mimic formal law by writing down their decisions and
registering and codifying their norms, usually as a means of enhancing the legiti-
macy of community proceedings and decisions.
The impact of state law and the extent of cooperation between state and indige-
nous authorities vary, depending on the geographic proximity, strength, and effec-
tiveness of formal legal institutions, the seriousness of the case, and the threat of state
repression of indigenous authorities. All of these variables help determine whether
indigenous law will compete with, complement, or substitute for state law. The most
common source of competition or conflict is the fact that indigenous and West-
ern justice institutions are based on distinct sets of cultural meanings and values
(Stavenhagen 1988, 102). A key normative conflict is the priority given in indigenous
law to restoring the harmony of the community, which may infringe on the rights of
individuals. There are crimes in Western law that are not considered such in some
indigenous communities, such as the marriage of minors, violence by husbands
against their wives, and the use of corporal punishment and the death penalty.
Conversely, there are transgressions in indigenous law that are not considered crimes
in Western law, such as gossip and religious dissent, which are sanctioned because
they disrupt the social order. Indigenous law may sanction the alleged use of super-
natural powers, which Western cultures do not recognize: the burning to death of
suspected witches is the classic example.
Collier (1998, 203–17) describes a case involving supernatural forces that oc-
curred in the community of Zinacantan, Chiapas, Mexico. According to local super-
stition, engineers routinely use human sacrifices to ensure that their constructions
endure, particularly highway bridges, which must include several human bodies or
body parts. Despite the absence of evidence against seven young men accused of
killing Indians and selling their bodies to engineers, they were found guilty and
required to admit their guilt and to pay the transportation costs for the authorities
and witnesses attending the trial in the municipal capital. Collier argues that the
indigenous judges achieved their goal of protecting the accused against mob vio-
lence and calming fears in the community. Nevertheless, from a Western liberal
perspective, the accused did not receive justice, since they were convicted of actions
that did not even occur.
Where indigenous law complements state law, it is typically due to efforts by
indigenous authorities to avoid conflict with state authorities. Indigenous authorities
the informal rule of law in latin america 255
have learned to select from their own and state systems according to their needs,
choosing the venue where they believe they will get the desired result. Indigenous
authorities use remittance to state authorities as an implicit threat should trans-
gressors wish to challenge their authority, just as the threat of state intervention
causes indigenous authorities to exercise their powers with discretion (Assies 2001,
87). Authorities are likely to use state procedures and remedies if they are not sure
how to proceed, when nonindigenous outsiders are involved, when their own law is
incomplete or unclear, or if they are worried about the penalties for breaking state
law (Orellana Halkyer 1998, 233). Indigenous authorities may serve as the first in-
stance of adjudication for grave matters, which may subsequently be referred to the
state system. They often bring the most serious offenses to state authorities, because
these usually are the result of conflicts that could not be reconciled within the
community or are instances of repeated violations. These are punished severely,
often with banishment, because they demonstrate a rejection of the indigenous
society’s norms. Going to outside authorities indicates that the community no longer
seeks to reintegrate the offender (M. Gómez 2000, 13–15).
Indigenous law at times substitutes for state law, particularly in remote rural areas
or in urban slums founded by migrants, where state institutions have failed to estab-
lish jurisdiction. Here, indigenous law operates without interference unless some
party to the dispute—or a concerned outsider—summons state authorities. In such
cases, relations may become competitive or accommodating, depending on the
inclination of the state and indigenous authorities to seek a mutually satisfactory
solution. For example, in lowland Bolivia, the Guaraní of Alto y Bajo Izozog burned
to death a suspected witch. Hearing of the incident, state authorities visited the
community to apprehend and punish ‘‘those responsible.’’ The entire community
claimed responsibility, so the police had no alternative but to leave, since the jail
would not accommodate hundreds of people. The incident prompted the Guaraní
to write down their laws. The written version, however, does not include execution as
a punishment for witchcraft, although anthropologists observe that the practice
continues.≥
Rondas Campesinas
the crime situation became less desperate, the rondas’ use of violence in the north-
ern villages studied by Starn decreased (Starn 1999, 89, 135; Degregori and Mariños
2000, 406; Ardito Vega 2001, 14).∑
Origins. The first rondas campesinas were formed in 1976 in the northern Peru-
vian department of Cajamarca to prevent and sanction the theft of livestock, which
was rampant and had increased with an economic crisis that pushed many peasant
families to the edge. The agrarian reform of the early 1970s had rapidly destroyed the
private hacienda authority structures prevailing in the sierra, without replacing them
with a system of effective public authority. Criminals had corrupted police and
judges, leaving campesinos to fend for themselves. In much of the northern sierra,
traditional cultural authority systems had long since broken down. Thus, the rondas
filled a vacuum of public and private authority. The absence of traditional, culturally
based authority structures, such as those persisting in the south, required the creation
of a totally new justice institution. The form it took was based less on ethnic tradi-
tions than on the hacienda guards that had been organized by landowners to protect
their property in an earlier era, on perceived police procedure, and on ronderos’
experience of military service. In the late 1970s, hundreds of Cajamarca commu-
nities formed rondas. The model also was disseminated to the department of Piura,
particularly after the El Niño of 1983 caused economic devastation that forced many
peasants to steal in order to survive. By the late 1980s, rondas had formed in the
departments of Amazonas, Lambayeque, and Ancash, involved more than thirty-four
hundred villages and more than four hundred thousand ronderos, and covered more
than sixty thousand square miles of the northern Andes (Starn 1999, 4, 18, 43, 109;
Degregori and Mariños 2000, 392–403).
Evolution. As crime diminished and the authority and prestige of the rondas
increased, at the insistence of villagers, rondas increasingly took on conflict resolu-
tion functions. Ronda assemblies were convened to administer justice in local crimi-
nal matters and to reduce ‘‘bickering and infighting’’ (Starn 1999, 113). There were
several reasons why the rondas were more effective than the state. First, they pos-
sessed the local knowledge and resources to investigate the facts of a case, whereas
local police and judges did not know the community and lacked the resources and
interest to make on-site investigations. Second, justice was speedy in ronda assem-
blies, which in the 1980s were held once a month, on average. Three-quarters of
cases heard in Tunnel Six, Piura, were resolved in one night, whereas in the state
system cases typically took three to four years to resolve. The threat of conveying a
transgressor to the police was usually enough to get that person to accept the author-
258 donna lee van cott
ity of the ronda, since ronda justice is preferable to languishing in jail for years before
trial. Third, trials were typically free, although Starn heard of isolated cases of ronda
presidents charging fees or demanding bribes. The state system required fees, nota-
ries, travel expenses, and lawyers, as well as time away from economic activities.
Finally, assemblies occurred in the disputants’ own community; it was not necessary
to travel to the foreign, intimidating environment of the city, where humiliation and
abuse were common. As the ronda assemblies became widely used in the northern
sierra—each village hearing approximately a hundred disputes each year—the local
police and courts saw a dramatic reduction in their caseload (Starn 1999, 123–38).
The rondas continued to evolve in the 1990s. By the end of the decade, they had
fallen into disuse in some areas: nightly patrols had been suspended and assemblies
were seldom held. This occurred in Piura, where state justice administration im-
proved dramatically in the 1990s and community leaders were tired of the unpaid,
thankless job of running the rondas (Starn 1999, 264). Elsewhere, in addition to
controlling crime, rondas act as political representatives before municipal authori-
ties and have become involved in important political issues, such as land titling. In
some areas they perform some of the roles assigned to municipal governments,
usually without conflict from the latter (Degregori and Mariños 2000, 407-9).∏
organized rival ronda federations. In 1986, APRA president Alan García promulgated
the Rondas Campesinas Law, which gave rondas legal standing to resist attacks from
police and the authority to protect individual and community property.
The army used the ronda model in the late 1980s and early 1990s to organize
southern campesinos against Sendero Luminoso (the Shining Path). By the mid-
1980s, the southern Andes, the center of fighting between the army and Sendero, had
become militarized. Those who suffered most were campesino communities, which
increasingly sought means to protect themselves from both sides. In the southern
highlands, traditional authorities and cultures were stronger than in the north and
these provided a resource for collective resistance. In fact, the first anti-Sendero
campesino rebellions occurred in response to attacks on indigenous community
authorities, particularly when these substituted young senderistas for older authori-
ties. To assist the campesinos in fighting Sendero, in 1991 the army distributed more
than ten thousand Winchester shotguns to the new rondas. But these antiinsurgency
rondas had an origin and purpose distinct from the ones described above, and they
were subordinate to existing forms of traditional authority (Starn 1992, 90).
Legal projects to define and constrain the rondas followed, and continue to this
day. Peru’s 1993 Constitution recognized their jurisdiction but, thus far, legal recog-
nition has been confined to protection of property and self-defense (Starn 1992, 105;
Degregori and Mariños 2000, 404). Degregori and Mariños (2000, 407) argue that
the Peruvian state has failed to take advantage of the potential of the rondas to fill the
huge gap in the availability of justice administration by formally linking them to the
local police, to local government, or to the system of justices of the peace. Perhaps
this is just as well, since the informality of the rondas continues to be an important
part of their identity and, perhaps, their strongest source of legitimacy and autonomy.
This situation may change as a result of Law 27908 (2003), which recognized the
jurisdiction of the rondas with respect to conflict resolution and called for local
public authorities to establish relations of coordination with them.
Juntas Vecinales
Most informal legal systems exist in rural, usually isolated communities, but in
Bolivia, migrants to squatter settlements on the outskirts of Cochabamba and La Paz
have constructed their own justice systems at the margins of state law. These systems
are deeply rooted in the Andean cultures of the rural sending communities, but their
written, internal rules incorporate Western norms, structures, and procedures.
This section is based on a 1997 Bolivian government study of four urban barrios
(Ministerio de Justicia 1997a, 1997b),π two in El Alto (on the outskirts of La Paz)
260 donna lee van cott
and two in the city of Cochabamba, that developed their own informal legal sys-
tems within the quasi-formal institution of the junta vecinal (neighborhood junta)
(1997b). More than a system of justice administration or dispute resolution, the
juntas provide a highly legitimate form of local government that represents the will
of the community. In all four cases there are two parallel, linked fora for conflict
resolution: the disciplinary tribunal, composed of a few of the most honest and
respected members of the community, and the direct action of the president and the
secretary of conflicts/justice. Although the duties of junta authorities are broad and
explicit—encompassing disputes over private property, abuse of public authority,
crime and delinquency, and domestic relations—the norms of justice administration
that they implement are not. The tribunal investigates complaints and offers a
solution or mediates an accord among the disputing parties. The solution is then
recorded in the Libro de Actas. In Villa Adela, El Alto, the tribunal conducts the
investigation and then forwards its report to the General Assembly of Neighbors,
which discusses the issues raised and chooses a sanction. The Federation of Juntas
Vecinales of the municipality might be asked to review the decision in cases where
the junta is unsure of itself (1997b, 65).
Frequently, people seek the direct intervention of the junta president or secretary
of conflicts. The written duties of these officials are typically to negotiate and find
consensus with respect to disputes over the boundaries of urban plots and the build-
ings upon them, to negotiate disputes over the use of land, to assist in drawing district
limits, and to serve as an intermediary in disputes among neighbors or between them
and another private or public entity. Most of these duties are actually illegal—juntas
do not have the legal authority to settle property issues, which are the purview of
municipal authorities. Nevertheless, the majority of disputes addressed by the secre-
tary of conflicts and the junta president are related to property disputes, which also
are the main source of intracommunity conflict in rural Aymara communities. Since
most urban property claims are not based on legal rights (claimants are squatters),
the juntas are the preferred venue (1997a, 23; 1997b, 39, 43).
Junta procedures incorporate Andean spirituality. In the Cochabamba barrio of
Alto Sebastián Pagador, it is common for the aggrieved party to consult a yatiri (seer)
to ascertain the truth of a matter, the person responsible (if unknown), or the appro-
priate sanction. In El Alto, aggrieved parties also may privately seek the advice of a
yatiri, and junta authorities will invoke the norms of their ancestors as they begin the
task of conflict resolution. Sanctioning takes place in the rural fashion, before a
crucifix and Bible, which are set upon a block of salt. The accused apologizes,
promises not to make the mistake again, and asks forgiveness. The sanctioning
phase, over which the oldest community member presides, most often draws on
the informal rule of law in latin america 261
Andean culture. The community elder’s ritual participation ‘‘seals the act with more
validity than a written paper’’ (1997b, 65; my translation). A fine is usually imposed
and registered in the Libro de Actas. The most common punishments are verbal
warnings (public or private), economic restitution and/or fines, and temporary sus-
pension of political rights. In the most serious cases or for repeated offenses—and this
happens only rarely—a person might be expelled from the community. Instead of
physical punishment there is a marked preference for resolving the problem with a
warning or nonpunitive agreement, or for the imposition of a fine (1997b, 63–64).
However, since punishments expected to elicit state condemnation or retribution
may not be entered in the juntas’ records, the government’s study may in fact
understate the incidence of corporal punishment.
Origins and Evolution. The origins of the juntas are similar to those of the rondas.
Communities in Bolivia formed the juntas to address the lack of state attention to
urgent community needs. Whereas the rondas reduced theft, the role of the juntas
was to defend, before the municipal government, the communities’ demands for
property and for access to urban infrastructure and public services to meet basic
housing, health, education, and transportation needs. Thus, in both cases, justice
administration was a subsidiary, albeit important, function at the time of origin.
Whereas the rondas were invented anew in the absence of traditional ethnic
authority structures, the juntas adapted contemporary rural Aymara and Quechua
cultural and authority institutions. Thus, the juntas may be considered either a new
informal institution or a relatively sudden transformation and transplantation of an
existing one. Although there is considerable variation, in many places urban barrios
are essentially mono-ethnic and may even be composed predominantly of migrants
from one province. Ethnic norms and authority structures are more predominant
in El Alto’s Aymara migrant communities, because their rural authority structures
(ayllus) are relatively strong. For example, the barrio of Alto Lima exhibits all the
traits of Aymara culture fused with the labor union experience of many Aymara ex-
miners. The urban justice system is quite similar to the rural Aymara justice system,
except that formal procedures are more strictly followed in order to lend a ‘‘greater
veneer of legality in an ‘illegal’ environment’’ (1997b, 69; my translation). In the
Quechua barrios around Cochabamba, by contrast, customs are as likely to be
adapted from campesino union norms and procedures, since most of the ayllus in the
Cochabamba valley have been destroyed and the region has experienced a greater
degree of mestizaje (cultural mixing) (1997b, 28).
In some respects, juntas vecinales are highly bureaucratized institutions. Authori-
ties are differentiated and ranked, and their functions and duties are carefully written
262 donna lee van cott
Relationship with the State. Relations between the juntas and municipal authori-
ties vary. In Villa Adela, El Alto, the junta has ‘‘fluid’’ relations with the police, and
these have improved in recent years. In Alto Lima Primera Sección, El Alto, where
there is no police presence despite the repeated entreaties of the junta, the junta
carries out police functions itself, using a system of whistles that alert the community
to a problem. If one or more of the parties to a dispute is not satisfied with the junta’s
decision, he or she may bring the dispute to the state authorities, but not before
requesting the permission of the junta. This is an unlikely event, since state justice
institutions are more costly and likely to be less helpful to culturally distinct mi-
grants. Conversely, junta authorities may decline to handle serious problems, such
as rape or murder, and refer them to the state (1997b, 58–71). An estimated four to
eight thousand juntas vecinales throughout Bolivia gained legal standing through
the 1994 Law of Popular Participation, which conferred collective participation
rights on campesino communities, indigenous communities, and juntas vecinales.
These new rights, however, are related to community planning and budgeting,
rather than justice administration. The juntas’ justice administration function is not
recognized in the 1994 constitutional language that officially sanctioned indigenous
customary law. Thus, like the rondas, the juntas vecinales are not formally permitted
to administer justice.
Table 12.1 summarizes the most important features of the three types of informal
justice institution discussed above. The commonality of norms, sanctions, duties,
and procedures among these types is attributable to the fact that all three systems are
based on the interplay of two features common to Latin America: long-standing
practices and cultural norms common to indigenous peoples in its rural areas, and
state procedures and norms that share similar features throughout the region. In all
three types of justice system, authorities have a broad mandate to resolve disputes
the informal rule of law in latin america 263
table 12.1
A Comparison of Informal Justice Administration Systems
Indigenous law Rondas campesinas Juntas vecinales
Reasons for Regulate social life; Protect against theft Replicate rural culture in
original perpetuate culture; relate urban space; present
emergence to colonial authorities collective demands to
local public authorities;
mediate property disputes
Current Preserve political and Deter and punish crime; Same as above
purposes territorial autonomy; resolve internal conflicts;
served complement or substitute provide more efficient,
for state justice system effective justice admin-
istration; represent
community to outsiders
Scope of Broad, including penal, Broad, including penal and Most commonly concerns
issues domestic, and religious domestic issues and access to land and public
considered issues, administrative law, dispute resolution services; also penal and
and dispute resolution domestic issues
Typical Mild corporal punishment, Mild, ritual whipping, Verbal warning, fines, loss
sanctions brief detention, eco- economic restitution, of property or political
imposed nomic restitution, com- community service, brief rights; most serious cases:
munity labor; more detention; more serious expulsion
serious offenses: expul- offenses: expulsion,
sion, severe corporal death, remit to state
punishment, death, remit justice system
to state system
Authorities Juridical authorities fused Village men elect Highly bureaucratized and
and organi- with religious and authorities; decisions hierarchical system of
zational political authorities, made by consensus of Western-style authorities,
structures chosen through community, with with differentiation of
community customs guidance of president functions
Relationship Varies, depending on status De jure subordination to Judicial functions not recog-
to formal of state recognition police and local courts; nized by the state; degree
system de facto autonomy, with of coordination varies
varying levels of
coordination with state
and to restore the harmony of the community in the absence of a more authoritative
and accessible institution. The relationship of these informal justice institutions to
the state varies according to the geographic and cultural distance of the state authori-
ties, as well as the propensity of community-state relations to be cooperative or
conflictual. The main difference among the systems is the extent to which traditional
indigenous cultural identities and norms permeate the provision of justice. Indige-
nous law tends to reflect modern-day interpretations of traditional indigenous cul-
ture and typically is practiced in the traditional rural community setting. Rondas
campesinas tend to adopt more features from state institutions to be used in tradi-
tional rural settings, whereas juntas vecinales bring traditional cultural practices to a
nontraditional, urban setting, which requires some adaptation to a new environment
and closer proximity to state institutions and formal politics.
264 donna lee van cott
The development of indigenous legal institutions entered a new era in the 1990s,
when Latin American constitutions began to formally recognize their public author-
ity and legal jurisdiction as part of a larger effort to recognize ethnic diversity and
collective rights for indigenous peoples. Bolivia, Colombia, Ecuador, Mexico, Nica-
ragua, Paraguay, Peru, and Venezuela explicitly recognized the multicultural or
multiethnic nature of societies in their revised constitutions. This recognition pro-
vides the normative framework for the recognition of legal pluralism. Bolivia, Co-
lombia, Ecuador, Peru, and Venezuela have extended the most constitutional or
statutory recognition to indigenous legal institutions, with Mexico also extending at
least formal recognition in the 1990 and 2001 constitutional reforms. In addition, the
constitutions of Guatemala, Nicaragua, Panama, and Paraguay make some refer-
ence to customary law (Assies 2001, 83) (see table 12.2).
There are three main reasons for this formal recognition. First, states responded
to intense pressure from indigenous organizations to recognize their collective rights
as peoples, following decades of activism and political organization by indigenous
peoples’ organizations. The demand for official recognition of indigenous law is an
integral part of the articulation of the demand of all indigenous movements in the
Americas: the demand for self-determination as peoples. This term encompasses the
right to self-government and the autonomy to freely develop their political, legal,
economic, social, and cultural institutions, while achieving full representation in
the state political system. In the face of centuries of state injustice directed toward
indigenous communities, the demand for autonomy in the administration of justice
is particularly important (Sieder 1999, 111).
Second, international norms for the treatment of indigenous peoples developed
in the 1980s require states to recognize indigenous law. Most important among them
is International Labor Organization Convention 169 (1989) on the rights of indige-
nous and tribal peoples. Among the broad set of rights that ILO Convention 169
codifies, it requires that states allow indigenous communities to conserve their legal
customs and institutions, provided that they do not violate fundamental rights as
defined by national or international law (art. 8, art. 9). Mexico (1990), Colombia
(1991), Bolivia (1991), Costa Rica (1993), Paraguay (1993), Peru (1994), Honduras
(1995), Guatemala (1996), Ecuador (1998), Argentina (1999), Brazil (2002), Do-
minica (2002), and Venezuela (2002) have ratified the convention, which also re-
quires that states adjust their national legislation to apply its norms, although the
table 12.2
State Recognition of Indigenous Law in Latin America
Bolivia: 1967 Constitution, reformed 1995, art. 171. The natural authorities of the indigenous and
campesino communities may exercise functions of administration and application of their own
norms as an alternative solution in conflicts, in conformity with their customs and procedures,
always providing that they are not contrary to the Constitution and the laws.
Brazil: 1988 Constitution and 1999 reforms, art. 231. The Indians’ social organization, customs,
languages, beliefs, and traditions are recognized, and the original rights over the lands that they
traditionally occupy.
Chile: Law 19,253(1993), art. 54. Custom shall be taken into account in judgments among Indians
belonging to the same ethnic group, it shall constitute law, provided that it is not incompatible with
the Political Constitution of the Republic. In penal law it shall be considered when it can serve as
an antecedent for the application of an exemption or extenuating circumstance.
Colombia: 1991 Constitution, art. 246. The authorities among the native peoples may exercise
judicial functions within their territorial areas in accordance with their own rules and procedures,
which must not be contrary to the Constitution and laws of the Republic.
Ecuador: 1998 Constitution, art. 191. The authorities of the indigenous peoples may exercise
functions of justice, applying their own norms and procedures for the solution of internal conflicts
in conformity with their customs or customary law, provided that they are not contrary to the
Constitution and the laws.
Guatemala: 1995 Accord on the Identity and Rights of the Indigenous Peoples, IV, E, 3. ‘‘In order
to strengthen the juridical security of the indigenous communities, the government promises to
promote before the legislative organism, with the participation of the indigenous organizations, the
development of legal norms recognizing the indigenous communities and the management of
their internal issues in accord with their customary norms, provided that these are not
incompatible with the fundamental rights defined by the national juridical system nor with
internationally recognized human rights’’ (Yrigoyen Fajardo 1999, 67–68).
Mexico: 1917 Constitution, mod. 2001, art. 2. This Constitution recognizes and guarantees the
right of the indigenous peoples and communities to self-determination and, in consequence, to the
autonomy to: I. Decide the internal forms of social, economic, political, and cultural organization.
II. To apply their own normative systems in the regulation and solution of their internal conflicts,
subjecting these to the general principles of this Constitution, respecting individual guarantees,
human rights, and in the manner relevant, the dignity and integrity of women.
Nicaragua: 1987 Constitution, art. 89. The communities of the Atlantic Coast have the right . . . to
have their own forms of social organization and to administer their internal issues according to their
traditions.
Panama: Recognized through ordinary legislation regulating the indigenous comarcas
(autonomous reserves), especially Law 16 of 1953, creating the first Kuna comarca.
Paraguay: 1992 Constitution, art. 63. The right of indigenous peoples to preserve and develop their
identity within their respective habitat is recognized and guaranteed. They have the right, as well,
to freely apply their systems of political, social, economic, cultural, and religious organization, as
well as the voluntary subjection to their customary norms for the regulation of their internal
lifestyle, provided that these do not violate the fundamental rights established in this Constitution.
Peru: 1993 Constitution, art. 149. The authorities of the Campesino and Native Communities,
with the support of the Rondas Campesinas, may exercise jurisdictional functions within their
territorial ambit in conformity with the customary law, always provided that they do not violate the
fundamental rights of the person.
Venezuela: 1999 Constitution, art. 260. The legitimate authorities of the indigenous peoples may
apply in their habitat instances of justice based in their ancestral traditions, and that only affect
their own members, according to their own norms and procedures, provided that they are not
contrary to the Constitution, to the law, and to public order.
Note: All language is my own translation from the original Spanish or Portuguese, except where noted.
266 donna lee van cott
rather than illegal, but not formal, since it remains beyond the regulatory reach of a
centrally controlled, normatively distinct legal system.
From the perspective of indigenous peoples, however, indigenous law is formal
law—binding, authoritative, and deeply rooted in indigenous norms and cultures,
even when some practices are of recent origin. Indigenous law constitutes, orders,
and preserves the sovereignty of indigenous society and forms of self-government,
just as state law constitutes, orders, and preserves the larger state and society. And this
is the perspective promoted by the new multicultural constitutions, which hold
indigenous law to be equal to state law by virtue of the equality of the cultures from
which both emanate and by virtue of the equal contribution of all cultures to the
nation. In addition, because of the underlying context of ethnic domination, indige-
nous law is an expression and an affirmation of indigenous peoples’ right to self-
determination and, thus, may be better described from the indigenous point of view
as a ‘‘counter-formal’’ institution.
achieve in practice, given the lack of knowledge about indigenous law among West-
ern legal scholars and judges and the complexity of the issues involved. These
difficulties can be divided into conflicts related to process, norms, and sanctions.
Conflicts of Norms. The conflict between a state’s liberal tradition, which privi-
leges individual rights, and indigenous peoples’ emphasis on collective rights has
prevented or delayed the writing of implementing legislation satisfactory to both
parties (Sierra 1998, 25). Often it is difficult for courts to determine what indigenous
norms are, since many Western norms and procedures have been incorporated into
indigenous cultures over the years. Another source of normative conflict is the use of
religion. In indigenous systems, religious beliefs may be central to the judgment,
investigation, and punishment of transgressions, since there is no distinction be-
tween religious and politico-juridical authority (Stavenhagen 1988, 101). Crimes
may be attributed to supernatural forces, which may conceal and perpetuate power
imbalances.
done this (Cabedo Mallol 1998, 7).∫ Although often contradictory, taken together the
Court’s rulings have established a broad scope for the exercise of indigenous law.
Because nearly all the issues that may be expected to arise have reached the Court in
the last decade, Colombia provides an interesting model for states trying to imple-
ment legal pluralism.
A 1994 case (Sentence T-254) established three important principles for linking
state and indigenous law. First, cultural traditions are to be respected to the extent
that those traditions have been preserved; that is, the less contact with and permea-
tion by Western culture, the greater the scope for cultural autonomy in the applica-
tion of special indigenous jurisdiction. Second, indigenous authorities must not in
their decisions or sanctions violate international human rights or fundamental con-
stitutional rights of a higher rank than cultural diversity. Third, indigenous law ranks
above ordinary civil law and ordinary legislation that does not protect fundamental
constitutional rights. The issue of conflicting views of due process was raised in a
1996 case (Sentence T-349) in which an indigenous defendant claimed his rights had
been violated because he had not been allowed to use an attorney before community
justice authorities and because the authorities had sentenced him to confinement in
a state jail, an unusual sentence for this community. The Court ruled against the
claimant, arguing that use of an attorney was not a community norm. In that case the
Court established an important rule: when limitations are necessary to defend con-
stitutional rights of a higher rank, such limitations must have the minimum possible
impact on indigenous autonomy and cultural integrity. The Court concluded that
these ‘‘intangible rights’’ are confined to ‘‘the right to life, the prohibition on slavery
and the prohibition of torture,’’ because there is an ‘‘intercultural consensus’’ on
these rights (Assies 2003, 4).
The most sensational indigenous jurisdiction case in Colombia occurred in
1996–97. Seven men were accused of being the ‘‘intellectual authors’’ of the murder
of the indigenous mayor of the mainly Páez (also called Nasa) town of Jambaló. The
Jambaló cabildo (community government) found the accused guilty and sentenced
them to whippings with a leather whip, expelled them from the community, and
stripped them of their political rights as Indians. The principal defendant took the
case to the municipal criminal court in Santander de Quilichao, which ruled that
the cabildo had denied the defendants the opportunity to defend themselves, that the
judges in the case were politically biased, and that the whipping constituted torture
and thus was illegal under international law. A higher court affirmed the lower court
ruling. The Páez Cabildo Association of the North then took the case to the Consti-
tutional Court, which ruled that whipping, although incorporated from Spanish
colonial culture, had become part of Páez culture and that its use inflicted no
the informal rule of law in latin america 271
permanent, serious harm (Sentence T-523⁄1997). Moreover, its object was not to cause
pain or humiliation but to purify the accused and thus facilitate his or her reintegra-
tion into the community (an argument that conflicts with the simultaneous expul-
sion of the accused in this case). Sánchez Botero deems this ruling to be ‘‘truly
paradigmatic from the hermeneutical point of view’’ (2000, 232), because it recog-
nized a broad scope of autonomy for indigenous authorities based on recognition of
the existence of non-Western symbolic orders that give meaning to non-Western
cultures.
In essence, Colombia’s Constitutional Court sustained Kymlicka’s ‘‘multicul-
tural’’ interpretation (1995, 167) of the responsibility of liberal states with respect to
violations of individual rights by illiberal national minorities: they must not impose
liberal values on groups that do not share those values, but must endeavor instead to
negotiate peacefully with illiberal national minorities, as they would with foreign
countries, except in cases of severe violations of human rights, such as slavery or
genocide. In fact, in Colombia, state agencies often intervene to negotiate a compro-
mise when an indigenous individual seeks protection against the illiberal rulings
of indigenous authorities. Sánchez Botero relates an enlightening example. The
mother and maternal uncle of a Wayúu girl who had been educated outside her
community instructed her to return to the community to be married following her
first menstruation. The girl filed a writ of protection with a court ‘‘to protect her
fundamental rights to education and free development of the person’’ (Sánchez
Botero 2000, 228). The court did not impose a solution; instead, representatives of
the girl and officials of the state Service for the Protection of the Minor negotiated
with the community authorities. The girl was allowed to stay in school and to decide
in five years whether she wished to return to the community, at which time her status
would be reevaluated. An exchange of animals occurred to compensate the family
for the loss of wealth that would have accrued from the girl’s marriage. This solution
restored the economic equilibrium of the tribe while protecting the minor from an
early forced marriage.
Another way of linking indigenous and state justice systems is through the cre-
ation of community justices of the peace, who work with indigenous authorities
while representing the state. Such projects are underway in Guatemala and Peru. In
Peru, justices of the peace sometimes incorporate local indigenous customs and
norms into their procedures and decisions or divide the work of justice administra-
tion with native or campesino authorities (Ardito Vega 2001, 10). Where local authori-
ties are effective and legitimate, justices may perform a mainly ceremonial or notar-
ial role and relegate most of the administration of justice to the rondas campesinas,
and even participate in ronda activities. In some cases, community authorities may
272 donna lee van cott
refer cases to the justices of the peace, while, in turn, justices may ask community
authorities to act as guarantors of their decisions or sanctions—for example, making
sure that the transgressor completes his community service—or to capture a commu-
nity member to be brought before the judge. In Guatemala, the United Nations is
sponsoring a project to create community justices of the peace to serve in Mayan
communities where there is no formal authority. These community justices of the
peace are not an organic part of Mayan social and political organization and, be-
cause the communities themselves did not select them, they may lack legitimacy.
They also challenge existing traditional authorities and indigenous mayors, and they
create a separation of political and juridical powers that does not exist in indigenous
culture (Murgas Armas 1999, 326-41). Nevertheless, the new community justices of
the peace have improved access to bilingual justice administration in Mayan com-
munities, while making it more culturally sensitive (Sieder 2003, 144–45).
conclusion
Where the state has extended formal recognition to informal institutions, they
take on a somewhat ambiguous status as autonomous informal institutions that are
protected from state regulation, subject to certain limitations. From the perspective
of indigenous peoples and the region’s most ‘‘multicultural’’ constitutions, informal
institutions derived from indigenous cultures may be considered ‘‘counter-formal’’
institutions, because of their legal equality and historical precedence with respect to
the state. But in most places, where there has been little or no legal recognition,
community justice institutions remain informal and relations may be competing
and conflictual.
What are the implications of the foregoing for the quality of democracy, par-
ticularly in the region’s most multiethnic countries? Latin American legal systems
that recognize indigenous law and other forms of informal justice administration
better correspond to their heterogeneous social reality. They are less discriminatory,
more flexible, and thus more just (Yrigoyen Fajardo 1999, 9). Improved access to
appropriate legal protection helps construct citizenship for society’s most excluded
groups, particularly in societies that have a long history of discrimination (Sieder
1998, 98). O’Donnell (1993) acknowledges this when he links ‘‘low-intensity citizen-
ship’’ to the inability of the state to establish the rule of law. Where states are too
weak to provide justice, informal institutions can substitute for state efforts and
thereby improve the quality of citizenship. Although formal recognition seems to be
necessary to avoid jurisdictional conflicts, it also seems necessary that informal in-
the informal rule of law in latin america 273
stitutions retain a good deal of the ‘‘informality’’ from which their autonomy and
authority are derived.
Many scholars observe that the autonomy and authority of informal justice institu-
tions enable them to serve as spaces of empowerment. Moore (1986) writes of ‘‘semi-
autonomous circles of power’’ or ‘‘semi-autonomous social fields,’’ while Nader (1980)
uses the term ‘‘subaltern counter-publics.’’ Assies (2001, 93) and Yrigoyen Fajardo
(1999, 41) view indigenous law as spaces of autonomy from which subaltern groups
can engage in an equitable intercultural dialogue with the state and the dominant
culture. Others argue that informal institutions may improve the quality of state law.
For example, Sierra (1998, 39) argues that indigenous peoples’ quest for autonomous
spheres of self-government can serve as a model of decentralized democratization for
Latin American societies, and Chambers (1997, 426–27) holds that recognizing
cultural diversity and human rights provides a source of legitimacy to states that have
lost the moral center previously derived from constitutional invocations of God and
Catholicism.
Legal pluralism advocates may overestimate the macrosocietal benefit of recog-
nizing informal justice institutions: first, because the deficiencies in democracy and
the rule of law in Latin America are so vast that many other reforms are needed to
achieve improvement; and second, because informal justice institutions have many
flaws that reflect the jealousies, power imbalances, cruelty, and ignorance that are
common to all human groups. Connecting informal and formal institutions, how-
ever, has provoked a constructive intercultural debate on the flaws of both, and their
increasing mutual accountability should draw attention to these flaws and the need
to rectify them.
Conclusion
actors enforce or comply with may be formal or informal. Instead of taking for
granted that formal rules are both effective and predominant, then, it is more accu-
rate to treat formal institutional predominance as one of several possible institutional
scenarios. Actors may operate in a context of strong formal and informal rules, as the
chapters on Chile show. Alternatively, weakly enforced formal rules may be trumped
by informal ones, as in the case of campaign finance in Brazil, norms of police
violence in Brazil, and candidate selection during the heyday of Mexico’s Party of
the Institutional Revolution (PRI). Finally, both formal and informal rules may be
weak, in that neither effectively guides actors’ behavior.≤ Scholars of Latin American
politics who assume away these latter institutional scenarios do so at great peril.
The chapters in this volume challenge widely held conceptions about informal
institutions in Latin America. For example, whereas informal institutions are often
viewed as deeply entrenched and slow to change, many are in fact quite dynamic. As
the Eisenstadt and Langston chapters on Mexico show (and as Carey and Siavelis
predict in the case of Chile), informal institutional change may occur quite sud-
denly. Moreover, as Van Cott’s chapter makes clear, even apparently long-standing
indigenous laws have, in many cases, been repeatedly modified and reinvented as
indigenous communities interacted with state authorities. An important lesson is
that informal institutions should not be treated as permanent fixtures in the political
landscape. Even where informal institutions prove highly robust (e.g., clientelism in
Honduras), it is essential to examine the sources of that stability.
The chapters also challenge the notion that informality is uniformly corrosive of
democratic institutions. Although corruption, clientelism, and patrimonialism re-
main widespread—and highly consequential—in Latin America, their pervasiveness
should not obscure the existence of informal institutions that strengthen the perfor-
mance or quality of democracy. These include unwritten power-sharing arrange-
ments that promote multiparty or executive-legislative cooperation in presidential
democracies (Siavelis; Carey and Siavelis), norms that sustain or enhance electoral
accountability (Stokes), and informal legal systems that deliver justice in communi-
ties where the rule of law is weak (Van Cott). To this list we might add numerous
informal institutions with ambiguous or double-edged effects. ‘‘Ghost coalitions’’
eroded transparency and accountability in the Ecuadorian legislature, but they also
enhanced governability (Mejía Acosta). Mexican concertacesiones undermined the
electoral process and circumvented the judiciary, but they also facilitated the resolu-
tion of postelectoral conflicts and helped the opposition National Action Party (PAN)
break down the PRI’s monopoly of power (Eisenstadt). And as Taylor-Robinson
argues, Honduran clientelism plays an important ‘‘substitutive’’ role by inducing
otherwise nationally oriented politicians to attend to local needs.
276 gretchen helmke and steven levitsky
The central themes of this volume travel beyond Latin America. Although we
cannot possibly do justice to the growing body of literature on informal institutions
in Africa, Asia, the Middle East, and postcommunist Eurasia, this section briefly
sketches out how the typology discussed in the introduction—complementary, ac-
commodating, competing, and substitutive informal institutions—might be applied
elsewhere.≥
As in Latin America, the bulk of the research on informal institutions in the
developing and postcommunist worlds has focused on competing informal institu-
tions, or those that directly subvert formal rules. Two types of competing informal
institution may be identified. The first is particularism—that is, particularistic norms
such as corruption, clientelism, and patrimonialism, which, as in Latin America, are
said to be both widespread and highly corrosive of new democratic institutions
throughout the developing world.∂ This is especially true in sub-Saharan Africa,
where neopatrimonialism, or ‘‘personalized rule . . . organized through clientelistic
networks of patronage, personal loyalty and coercion’’ (Lindberg 2003, 123), is said to
be a defining characteristic of most emerging multiparty regimes (Bratton and van
de Walle 1994). In Ghana, for example, pervasive neopatrimonialism is said to have
undermined the formal checks on executive power outlined in the 1992 Constitution
(Sandbrook and Oelbaum 1999), and the clientelistic practice of ‘‘chop,’’ in which
members of parliament are expected to distribute particularistic resources to sup-
porters, is seen to ‘‘threaten the very heart of parliament as an institution’’ (Lindberg
2003, 136).
Particularism is also pervasive in postcommunist Eurasia. According to Joszef
Borocz, clientelism, graft, and other particularistic practices are so widespread in
Central Europe ‘‘that conducting any business, economic or otherwise, is virtually
impossible without bowing, or even succumbing, to’’ such practices (2000, 348).∑ In
the countries of the former Soviet Union, ‘‘particularistic rules and norms’’ routinely
trump the formal rule of law, and, in extreme cases, formal rules ‘‘serve as merely a
‘façade’ covering informal dominance’’ (Gel’man 2003, 92–93).∏ In Central Asia, for
example, clan politics so subverted formal institutions during the 1990s that ‘‘infor-
mal mechanisms of network-controlled exchange and norms . . . became the rules
of the game’’ (Collins 2002b, 23). Consequently, the formal political regimes that
emerged in the region following the collapse of the Soviet Union became ‘‘de-
institutionalized and inconsequential’’ (Collins 2002b, 30).
conclusion 277
enforce the rule of law or provide even basic public goods across the national
territory (O’Donnell 1993). In many countries, much of the national territory—
especially rural areas—is characterized by the ‘‘unrule of law’’ (O’Donnell 1993,
1999c). The absence of effective state enforcement of formal rules and laws allows for
the survival of indigenous institutions, as well as the emergence and persistence
of particularistic institutions such as clientelism, corruption, and mafias and clan
networks.
Furthermore, many developing countries are characterized by severe socioeco-
nomic, ethnic, and/or regional stratification, which makes it more likely that certain
groups—for example, ethnic minorities—will be denied access to (or protection
from) formal state institutions. As Van Cott’s chapter suggests, such exclusion may
lead these groups to maintain or (re)create substitutive informal institutions at the
margins of the state legal system.
A second observation concerns the flip side of the coin: the relative absence in the
developing world of underlying informal institutions that reinforce formal ones.
Although the ineffectiveness of formal institutions is usually attributed to the pres-
ence of subversive informal institutions, such an outcome may also be rooted in the
absence of complementary ones. Functioning formal institutions do not operate in a
vacuum. Rather, they tend to be embedded in shared norms and expectations that
facilitate or encourage compliance with them (North et al. 2000; Galvan 2004;
Stokes, this volume). As Dennis Galvan (2004, 18–19) has noted, liberal democratic
constitutions work effectively when they are embedded in underlying norms such as
gracious losing and a strict public-private boundary. Similarly, as Stokes argues in
her chapter, elections serve as mechanisms of vertical accountability only where
there exists a shared expectation that voters will follow retrospective decision rules.
In the absence of such underlying norms, even systematically enforced formal
institutions may not work well. Consider presidentialism. Juan Linz (1990) has
argued that the constitutional arrangements of presidentialism (separately elected
executives and legislatures, fixed presidential terms in office) generate political be-
havior (such as zero-sum politics and executive-legislative conflict) that can destabi-
lize new democracies. Indeed, the United States is the only country in the world in
which presidential democracy did not break down during the twentieth century.
This outcome is explained, in part, by the array of ‘‘paraconstitutional’’ norms and
practices in which U.S. presidentialism has long been embedded (Riggs 1988).
Indeed, as Stokes (2003, 5–6) has noted, the authors of the Federalist Papers were
keenly aware of how their proposed constitutional arrangements would interact with
existing social norms. In Latin America, by contrast, postcolonial elites essentially
conclusion 281
tions in Latin America and elsewhere. These include questions of why and how
informal institutions are created, the sources of informal institutional stability and
change, and the conditions under which informal institutions are formalized. They
also include the dynamics of interaction between formal and informal rules: how
informal institutions contribute to formal institutional stability or change—and vice
versa. Compared with the study of formal institutions, the ‘‘learning curve’’ in infor-
mal institutional analysis—particularly in terms of establishing their existence—is
fairly steep. However, because the identification of informal institutions requires
that scholars look closely at the actors and mechanisms that sustain them, the very
process of identification may go a long way toward identifying sources of continuity
and change.
Although political scientists have long recognized that informal institutions mat-
ter, systematic research on the subject remains at an incipient stage. As this volume
has shown, advances in these areas are likely to take place on a variety of method-
ological fronts, ranging from abstract formal modeling to fine-grained historical or
ethnographic studies. Theoretical insights into the origins, dynamics, and effects of
informal political institutions will likely draw from a range of disciplines, including
anthropology, economics, law, and sociology. It is therefore essential to maintain a
broad and pluralistic research agenda that encourages fertilization across disciplines,
theoretical traditions, and methods. It is this spirit that gave rise to this volume—and
to which we hope the volume will contribute.
afterword
guillermo o’donnell
tually for the condemnation of those informal institutions that most egregiously
transgress them. In this convoluted and admittedly less than optimal way, democra-
tizing discourses and demands are anchored in and by the formal rules that even very
informal rulers cannot cease to evoke.≤ This happens repeatedly under democratic
regimes, and this is one of their main, if seldom acknowledged, advantages in rela-
tion to other kinds of regimes—democracy, even if ‘‘formal’’ or ‘‘electoral,’’ has some
discursive and ritual requirements that not even scarcely democratic leaders can
fully bypass.
It seems to me that these strategic shifts from and to formal and informal are the
stuff of politics, particularly where regimes are poorly institutionalized. Thus, the
typology proposed in the introduction might be complemented by a focus on how
and why some actors adapt, skillfully and recurrently, to codes embodying various
sets of rules, and on how and why the relevant others in the respective contexts
respond to these shifts. In this perspective, the focus would be not so much on
the institutions themselves (formal and/or informal) but on how actors ‘‘within’’
these institutions behave, and on how they strategically shift according to varying
contexts.≥
Now I take a step back in these reflections. In my early articles on this matter, I
expressed concern with the excessive formalism of many of the works in the first
wave of institutionalist studies on Latin America (and, in fact, elsewhere). I was (and
am) persuaded that, as the editors note and the chapters in this volume amply show,
‘‘analyses of democratic institutions that focus exclusively on formal rules thus risk
missing much of what shapes and constrains political behavior, which can yield an
incomplete—if not wholly inaccurate—picture of how politics works.’’ My concern
was centered on competing informal institutions: those that through various sets of
informal rules (clientelism, nepotism, corruption, and others) guide behaviors that
are inconsistent with, if not opposed to, the formal rules of democracy. Seen from
this angle, it seems to me that the first question to be asked in these matters is, to what
extent do the formal rules predict actual, observed behavior (and in some cases
omissions)? In some cities, as we saw, the formal rules of transit are very poor
predictors in most situations, albeit pretty good ones in others. Based on an ad-
mittedly nonscientific observation of politicians, congresspersons, and some presi-
dents in Argentina and Brazil, I believed that this was also true of them. Several
chapters in this volume support, from various angles, this view.
On the other hand, it is trivial to assert that informal rules and institutions exist
everywhere. The matter is, to what extent and in what areas (the political one, in the
case of this volume and my articles) informal rules and institutions actually govern
behavior, and to what extent and in what areas such behavior is inconsistent with, or
288 guillermo o’donnell
opposed to, the formal rules of democracy. As several contributors to this volume
note,∂ these deviations may not necessarily be harmful to the workings of a given
political regime, but if they are importantly influential, then something significant
has been found that cannot be accounted for by views that presuppose the pre-
dominance of formal rules. Of course, studying these matters is empirically difficult,
but, as this volume illustrates, various methods can deal with them in quite satisfac-
tory ways.
Still, these matters need a bit more conceptual precision, for which the typology
proposed by Helmke and Levitsky is apposite. One matter, of course, is what contem-
porary political scientists know about the role of complementary and accommodating
institutions in facilitating or partially correcting the workings of formal institutions.∑
In fact, this role was detected earlier, and abundantly discussed, by sociologists, so
much so that the contemporary sociology of the organization was born precisely with
this discovery.∏ Thus, the existence of these kinds of informal institution does not
entail a valid objection to the study of other kinds, including competing and substitu-
tive ones. One is free to study whatever kind of institution seems interesting, al-
though it may not come as a surprise that many students of democratization choose
the latter types, precisely because they are critically interested in how to improve the
workings of these highly informalized democracies.
This disagreement points to the need for another bit of conceptual precision,
which is my third suggestion. The editors take good care in stipulating what informal
institutions are not. Perhaps it may be added, in more detail, that such institutions
are constituted by sets of informal rules. These rules, as such, have at least three
characteristics: (1) they entail some kind of sanction in case of being violated or
grossly ignored, even though there may not be a predesignated actor to apply the
sanction; (2) they are common knowledge in the relevant context of interaction; and
(3) they prescribe some kind of action (or, in some cases, omission) on the part of the
subjects acting in the relevant context. Thus, in addition to the differentiations of
informal institutions vis-à-vis other phenomena that the editors correctly stipulate, it
would be useful to add that informal rules are not simply expectations, even though,
obviously, they generate them. This may dispel confusions such as the assertion
that voters’ beliefs that candidates will honor their campaign promises are informal
rulesπ —these beliefs are expectations, not rules in any reasonable definition of the
term of which I am aware.
Following the useful typology proposed in the introduction, we may note that
complementary, accommodating, and competing informal institutions are all in some
kind of relationship with formal ones. As Helmke and Levitsky note, in many cases
the informal may be endogenous to the formal, in the sense that they originate in
afterword 289
and through the workings of the latter. Yet an intriguing possibility is the reverse.
Particularistic practices in informal institutions have been central to the trade of
many Latin American politicians before and during authoritarian regimes.∫ Thus,
even though this does not make the formal institutions of democracy endogenous to
the informal ones, they may have been, as it were, plunged into a deep sea of
preexisting informal rules and institutions. This may account for the resilience of
these informalities in spite of so many efforts (and money from international institu-
tions) to ‘‘modernize’’ parties, congress, the judiciary, and the like. This might also be
helpful in finding clues for reducing the weight of those informalities. One way or
the other, we do not know much about the mechanisms through which each kind of
institution relates to any other, and even less about the consequences of the temporal
sequences in which they are effected,Ω but one useful characteristic of this volume is
to open the way for this kind of inquiry.
The lack of endogeneity of informal institutions in relation to formal ones is also
illustrated, albeit from a different angle, by Van Cott’s chapter. As this author shows,
indigenous legal systems, although influenced by the state and sometimes forced to
partially adapt to it, have mostly developed on the basis of a dynamic of their own.
Furthermore, I doubt that the institutions that Van Cott discusses should be dubbed
informal. Truly, most of their rules are not written (although Van Cott shows an
increasing movement toward written law), but these indigenous legal systems and
institutions are quite formally effected, including publicly appointed and legiti-
mized authorities, detailed procedures, elaborate rituals, regularized sanctions, and
the like. Consequently, these institutions are not only public knowledge but also
transparent, not just for those directly involved in the relevant context of interaction
but also for outsiders. In any event, this chapter suggests that in some cases there may
be a gray zone between formal and informal institutions—as the use of any dichot-
omy in the social sciences would unfailingly find.
As springs, I hope, from the preceding discussion, in addition to its obvious
intrinsic merits this volume is a welcome challenge. It would have been nice if
implanting new formal rules and institutions had sufficed for producing the ‘‘consol-
idation’’ of democracy, in Latin America and elsewhere. The problem is that, as this
volume abundantly illustrates, those rules and institutions often are poor predictors
of behavior. This defies us, not to neglect formal rules and institutions, but to
undertake careful studies of, as the editors propose, the interactions between the
formal and the informal and, as I suggest here, the ways in which actors strategically
navigate formal and informal contexts. The progress of social knowledge does not
admit the shortcuts implied by simplistic formalists, importers of ready-made mod-
els, or hurried consultants.
This page intentionally left blank
Notes
introduction
The author thanks John Carey, Gretchen Helmke, Steven Levitsky, Maria Victoria Murillo,
and Ignacio Walker for their many comments and suggestions, which improved this chapter.
1. The Spanish terms are used here because they are used in the press and in everyday
parlance in Chile—attesting to their centrality in Chilean political life.
2. In dozens of interviews with members of the legislative and executive branches during
the Aylwin government, few interviewees failed to cite the ‘‘special conditions’’ and/or ‘‘special
circumstances’’ of the transition that affected their behavior (see Siavelis 2000, 51–52).
294 notes to pages 40 – 67
3. Cuoteo carries a negative connotation, and an implicit suggestion that the quota is
reached behind closed doors.
4. Exceptions include Deheza (1997) and Amorim Neto (2000).
5. Usually said to include SEGPRES (Ministry of the General Presidency) and the minis-
tries of Interior, Government, Defense, and Foreign Relations.
6. For a small sampling of the controversy surrounding the cuoteo in ministerial appoint-
ments and some of its negative consequences, see Qué Pasa (2000), and El Mercurio de
Valparaíso (2003).
7. While this is an account from negotiations on the right, as is the example cited in the text
below, the electoral system provides these same incentives and logic of pairings across the
political spectrum; Andrés Allamand, interview, Washington, DC, July 13, 1998. On the signifi-
cance of the role of small parties in general, see El Mercurio (1993b).
8. For a complete discussion of the rules governing candidate selection, and the incentives
they generate depending on party size, see Siavelis (2002b).
9. For some of the many examples of this phenomenon, see El Mercurio (1993a) and La
Epoca (1993a, 1993b).
10. Ignacio Walker, telephone interview, March 3, 2004.
11. Carlos Carmona, interview, Santiago, Chile, April 23, 1993.
12. Ignacio Walker, telephone interview, March 4, 2004.
13. Cesar Ladrón de Guevara, interview, Santiago, Chile, May 13, 1999.
14. Correa quoted by Ignacio Walker, interview, March 3, 2004.
15. Concertación leaders also often engaged in direct negotiations with influential actors on
the left, primarily trade unions, when drafting and advocating the passage of controversial
legislation, pointing to a consistent pattern of pact-making even when attempting to satisfy the
coalition’s natural constituencies.
1. State coffers may directly provide jobs or even public services that can be distributed—for
example, enrollment in a public school or access to a public health clinic offered to supporters.
Others have suggested that politicians deliver pork projects in exchange for kickbacks from
construction firms (Samuels 2001a, 2001b).
2. Note that in many clientelistic systems, parties play very important roles in politics. But
in Brazil, budget power is vested in strong executives, so parties’ role in clientelistic electoral
markets is minimal.
3. The interviews in Piauí were conducted in April 1999.
4. The interviews in São Paulo were conducted between December 1998 and March 1999.
5. Deputy Leal Junior, interview, Piauí State Legislative Assembly, May 1999; Leal Junior
quoted in Ana Cláudia Coelho, ‘‘ ‘Oposição se faz com dois, três, cinco, ou dez,’ afirma Leal
Júnior,’’ Meio Norte, March 5, 2000.
6. Deputy Tonin, interview, and Deputy Dias, interview, São Paulo State Legislative
Assembly, January 1999.
notes to pages 69 – 73 295
The author thanks Daniel Brinks, Max Cameron, Michael Coppedge, Fran Hagopian,
Gretchen Helmke, Steven Levitsky, and Michelle Taylor-Robinson for their insightful com-
ments and valuable feedback
1. For a broader review of alternative theories explaining the success and failure of market-
oriented reforms in Ecuador, see Mejía Acosta (2004, 3–9).
2. In 1999 alone, Ecuador’s economy was severely affected by El Niño–related floodings
(with an estimated loss of 13% of 1998 gross domestic product), a drop in international oil prices
(from $20.45 a barrel in 1996 to $6.95 in 1998), and the contagion effect from the Russian crisis
(drying up of international credit, soaring interest rates).
3. In the mid-1990s, several reforms were passed to abolish term limits, eliminate midterm
elections, and adopt a peculiar open-list proportional representation system that allowed for
personalized voting. For a more complete review of the implications of institutional reform for
coalition formation, see Mejía Acosta (2004).
4. Presidents have exclusive authority to initiate the budgetary process and other economic-
related legislation, to contract and acquire foreign debt, and to grant the required licenses and
contracts for the administration of the public sector (Political Constitution of Ecuador, art.
171). They also have some judicial prerogatives to pardon and reduce sentences and pass
controversial legislation by plebiscite.
5. The data reported are net public approval rates (all favorable minus unfavorable) for the
president. Support for congressional performance (not reported here) consistently remains in
the negative numbers as well.
6. Wilfrido Lucero, ID congressman, interview, Quito, July 19, 1999.
7. Jamil Mahuad Witt, former DP (Popular Democracy) party member and former presi-
dent, interview, Cambridge, MA, July 9, 2002. According to Mahuad, one party leader and
former president of Ecuador explained that criticizing him (Mahuad) was a matter of public
reputation, since ‘‘he could not [seem to] be less tough on the president than political leaders
from other parties.’’
8. Interviews with members of Ecuadorian National Congress, Quito, July 1999.
9. Some interesting examples are camisetazo, shameless change of political party with rent-
seeking purposes; colaboracionismo, an accusation made against parties that accept govern-
ment appointments without belonging to the president’s party; liborios, legislators subservient
to the government; teta, compensations received for government collaboration; and chuchu-
mecos, last-minute government allies.
10. The first bill included some basic reforms needed to implement dollarization in Ec-
uador; it was approved at the end of February, only fourteen days after its initiation (for a full
report, see Economist Intelligence Unit 2000, 16). In mid-March, Noboa proposed additional
amendments demanded by the International Monetary Fund in order to sign the stand-by
agreement with Ecuador, and although the congressional majority did not fully approve of
those reforms, it accepted the government’s partial veto.
11. The index is an aggregate measure of commercial, financial, capital accounts, privatiza-
296 notes to pages 73 – 76
tion, and tax reforms between 1970 and 1995; it ranges from 0 (complete state intervention in
these policy areas) to 1 (economic liberalization) (Morley et al. 1999).
12. For an interesting exception to this trend, see Gibson and Calvo (2000); they claim that
clientelistic practices were efficiently used to produce cheap legislative coalitions for eco-
nomic reform in Argentina.
13. Mainwaring (1999, 180) rightly points out that neither clientelism nor patrimonialism is
coterminous with corruption. Obtaining state resources for a specific region or distributing
government positions to political partners can be legitimate practices, or not proscribed by the
constitution. However, operating in poorly institutionalized polyarchies with eroding legal
authority and no (vertical and horizontal) links of accountability, clientelistic exchanges cre-
ated huge temptations for corruption.
14. Alexandra Vela, DP congresswoman, interview, Quito, July 11, 2001.
15. This is, however, not exclusive of Latin American presidential systems, as other scholars
have illustrated the roles of clientelistic exchanges in parliamentary systems as well.
16. Another concealed strategy to ‘‘support without supporting’’ was applied by the ID,
when it enabled Esparza’s election by annulling its vote, thus adding support to the win-
ning candidate according to internal congressional rules. ‘‘La convergencia al revés,’’ Vistazo
(Guayaquil), August 26, 1983.
17. Quoted in Mills 1984. After his term in congress, Gary Esparza was given control over
the Customs Administration, presumably as compensation for his role in facilitating legislative
cooperation with the Hurtado government.
18. Diario Hoy (Quito), January 24, 2001.
19. According to the Internal Legislative Rules, any legislator—with the support of ten
colleagues—can request a roll-call vote (Reglamento Interno de la Función Legislativa, chap.
IV). This is an unusual procedure used for controversial votes. A more common voting mecha-
nism is the simple vote, in which legislators raise hands to support a motion on the floor.
20. Interview with anonymous congressman from the independent MIN, Quito, April 11,
2001. The congressman added: ‘‘It’s like former Mexican President Echeverría used to say to
his chauffeur, ‘always signal to the left when you are going to turn to the right.’ ’’
21. In Ecuador, for instance, a newly appointed health or public works minister would, in
turn, have the authority to appoint provincial directors, allocate government contracts, and
include her own political cronies on the government payroll.
22. After the 1998 constitutional reforms, congressional authorities were elected every two
years, and there was an explicit provision to select the president of Congress from among the
two largest parties.
23. Agendas could include (1) administration of justice through the joint presidential and
party nomination of Supreme Court judges (between 1979 and 1996) or the nomination of
Constitutional Tribunal members (since 1997); (2) election oversight through the designation
of the members of the directorate of the Supreme Electoral Tribunal and those of its provincial
branches; (3) control and oversight of government administration through the appointment of
such authorities as the attorney general, the general comptroller, and the superintendents of
banking and telecommunications, as well as various directors of state-owned enterprises.
24. Pro-government coalitions are those composed of parties that vote for the executive
branch’s candidate in annual congressional leadership elections.
notes to pages 76 – 80 297
25. Roughly speaking, Ecuadorian presidents had the authority to appoint at least two
hundred or so government officials, including cabinet ministers and their subcabinet secre-
taries, for all ministries.
26. Similarly, in her work on coalition making in Italy, Mershon (1996, 538) argues that
politicians were able to ‘‘raise’’ the payoffs of coalitions by expanding—among other things—
the number of subcabinet positions available to coalition partners.
27. Interview with former Ecuadorian president, held anonymous at his request, Quito,
April 2001.
28. Through these positions, political parties gain access to significant provincial-level or
area-specific resources, grant concessions and licenses to diverse interest groups, negotiate the
channeling of funding for a locality, appoint and remove lower government bureaucrats, and
use their diplomatic status to further trade, tourism, and cultural policies.
29. The increment was especially significant in the most flexible spending accounts: trans-
ferences, (local) allocations, general spending, and public works (Araujo 1998, 145).
30. A 1996 survey of Ecuadorian legislators confirms this clientelistic connection. When
asked how important it was for them to obtain resources for their communities, more than 90
percent of legislators from center or right-wing parties, and more than 80 percent from center-
left parties, responded that it was ‘‘very important.’’
31. It seems that President Noboa bought individual votes to push a fiscal reform through
Congress in 2001. On May 4, he blocked a congressional override and passed a 2 percent VAT
(value-added tax) increase, with the help of several legislators who failed to vote or voted
against their party leadership. In subsequent weeks, the media reported the nature of payoffs
received by defecting legislators: Reynaldo Yanchapaxi (a veteran DP legislator from Coto-
paxi) obtained better roads for his province; Raúl Andrade (a legislator from the Ecuadorian
Roldosista Party [PRE]) obtained a fast-track credit for his native province, Manabí; Fulton
Serrano (PRE) bargained to get the governorship of his province for his son and directorship of
the Health Ministry for another family member. ‘‘Noboa relies on political quotas,’’ Diario El
Comercio (Quito), July 8, 2001.
32. Alexandra Vela, DP congresswoman, interview, Quito, July 1999.
33. On average, only one in ten legislators in Ecuador abandoned their parties in a
given year, despite the nonexistent or ineffectual punishments for party defection. By con-
trast, Brazilian legislators, operating in an equally fragmented and polarized legislature,
switched parties two or three times more often than their Ecuadorian counterparts (Main-
waring 1999).
34. See Sánchez-Parga (1998, 84, 101, 106).
35. Detailed information about the party affiliation of such government officials is mostly
nonexistent or incomplete.
36. After 1998, legal provisions were set in place to raise the threshold required to impeach
and censure members of the cabinet.
37. I thank Michel Rowland for sharing with me his cabinet survival dataset (1979–2002).
38. Perhaps a classic example is that of PSC Deputy León Febres Cordero (1979–84), who
launched several cabinet impeachment initiatives before earning an electoral reputation that
helped him gain the presidency in 1984.
39. In this ‘‘letter of commitment,’’ legislators also requested that the president pass—or not
298 notes to pages 80 – 92
block—necessary reforms to allow immediate legislative reelection. Diario Hoy (Quito), Au-
gust 13, 1993 (cited in Landau 2001, 128).
40. In the case of Montero, switching sides did not affect the electoral support in his small
southern province of Loja, as he returned to Congress under a different party label (CFP) in
1996.
41. ‘‘Government-Congress Relationship Sours,’’ Weekly Analysis, August 19, 1993, 6 (cited
in Landau 2001, 45).
42. Revista Vistazo (Quito), May 19, 1994 (cited in Hey and Klak 1999, 78).
43. When the PSC lost the election for president of Congress to Carlos Vallejo (ID) in
August 1992 (Burbano de Lara and Rowland 1998), a government alliance helped it regain
control of the courts and Electoral Tribunals until 1994 (Congreso Nacional del Ecuador
1995).
44. Party switching rates reached a record high of nearly 25 percent of legislators during this
period. At some point, the ‘‘group of independents’’ became the second largest contingent in
Congress after the PSC, giving them great bargaining leverage vis-à-vis the executive (Mejía
Acosta 2003).
45. Diario Hoy (Quito), August 13, 1993 (cited in Landau 2001, 128).
46. Alberto Dahik, telephone interview with David Landau, January 30, 2001 (cited in
Landau 2001, 47).
47. Dahik interview (by Vivanco) and Febres Cordero’s comments on Dahik in Diario Hoy
(Quito), July 7, 1995; Castello’s comments in Congreso Nacional del Ecuador (1995).
48. In Ecuador, some corruption watchdogs, such as the Civic Committee against Corrup-
tion, have been endowed with some jurisdictional powers, and important legislation has
been approved to mandate disclosure of campaign financing, government contracting, and
access to government information. Finally, several civic organizations promoting transparency
and accountability have appeared in the Ecuadorian context, but none that act as legislative
watchdogs.
49. The first two reforms are included in the 1998 Constitution; the third is codified in the
1998 Legislative Ethics Code.
50. Indeed, since the constitutional reform, not a single elected president has successfully
completed his mandate.
51. Ramiro Rivera, DP congressman, interview, Quito, July 1999.
1. A worthy question is the extent to which other informal institutions emerge in contexts
where no formal institution whatsoever regulates the question at hand.
2. I am excluding, for tractability, the possibility that a politician provides services in
exchange for expected financial support at some undetermined later time.
3. This argument holds independent of whether corruption exists and/or whether those
with money also attempt to influence nonelected officials. Only if elected officials are mere
window dressing will access-seekers not see them as potential investment targets.
notes to pages 92 – 111 299
4. Of course, recent scandals in Mexico, Germany, and elsewhere point to the likelihood
that even where campaigns are publicly funded, politicians will seek additional sources of
money illicitly, whether for campaigns or for personal enrichment.
5. This hypothesis holds independent of whether suppliers believe that it matters who wins;
only if contributors believe candidates are interchangeable and that politicians cannot affect
their interests will they not supply funds.
6. These factors are too numerous to explore here, but include the costs of newspaper and
television advertisements, other publicity efforts such as the manufacturing of pamphlets,
banners, and tee-shirts, transportation costs, get-out-the-vote efforts, and the costs of running a
campaign organization. These costs vary with local conditions.
7. I discuss elsewhere the reliability and validity of the data (Samuels 2001a, 2001b, 2001c,
2002). Cox and Thies, defending the use of campaign finance data from Japan, concluded that
‘‘if these data have been fabricated, they have been fabricated so as to preserve a number of
expected correlations and even to fit the theories of political scientists—which does not seem
too likely’’ (2000, 45). The same holds for the Brazilian data, justifying their use.
8. For 1994 and 1998, candidates were required to report only contributions, not expendi-
tures. I assume that candidates spent all they raised, because by law candidates are required to
hand over all ‘‘leftover’’ funds to their national party organization.
9. Television and radio time is free, distributed according to parties’ proportion of seats in
the lower chamber of Congress.
10. In-depth field research might find that family relations extend to in-laws, cousins, and
others who do not carry the candidate’s family name, and might also reveal that corporate
contributions come from firms for which the candidate had previously worked or which he
owns, or in which he holds a large stockholding share.
1. What representation means is a matter of debate: it has been defined as the act of
representing the interests of the electorate as delegates or as trustees, or of granting all groups a
place at the table where policy is deliberated, among other formulations (Pitkin 1967; Mans-
bridge 2003).
2. This would mean that patron-client relations are the foundation for the rootedness of
parties in society that Mainwaring and Scully (1995) offer as one characteristic of an institu-
tionalized party system.
3. This literature summary draws primarily on the theory presented by Carey and Shugart
(1995), but also on Lancaster (1986), Carey (1996), Mainwaring and Shugart (1997), and
Shugart (2001).
4. See Rothstein (1979) for a discussion of how Mexico’s PRI (Party of the Institutional
Revolution) used patron-client relations to allocate scarce resources for local infrastructure
development.
5. Despite its name, the Liberal Party has always been relatively conservative. Both tradi-
tional parties defend traditional elite interests, and neither is ideological.
6. Interview with PLH deputy, Tegucigalpa, Honduras, July 30, 1997.
300 notes to pages 112 – 114
7. In the 2002–5 Congress, the three small parties won twelve seats, and for the first time the
president’s party lacks an absolute majority in Congress (his National Party holds 61 of 128
seats).
8. Until 1993, municipal elections were also fused with the presidential and congressional
elections.
9. I could not verify these claims, but they are an indication that local support matters to
aspiring politicians.
10. For a detailed explanation of the coding scheme for bill targets, see Taylor-Robinson
and Diaz (1999). For bills with multiple authors, the bill is counted in the legislative record of
each author, because records in the Honduran Congress do not indicate a ‘‘primary’’ author or
‘‘cosponsors.’’ Multiauthored bills are very rare in Honduras and typically have only two
authors, so this will have little impact on the analysis.
11. It is hard for an individual legislator to claim credit for curbing inflation or for a national
highway program (Mayhew 1974, 59–60). Big projects require the support of many members,
so credit must be shared. There is also an informational challenge to claiming credit for such
projects. As Mayhew writes: ‘‘For typical voters Capitol Hill is a distant and mysterious place;
few have anything like a working knowledge of its maneuverings. Hence there is no easy way of
knowing whether a congressman is staking a valid claim or not . . . [and for] a voter lacking an
easy way to sort out valid from invalid claims the sensible recourse is skepticism’’ (1974, 60). If
lack of information makes it difficult for U.S. congressional representatives to claim credit for
their legislative activities, it is likely even more difficult for Honduran deputies, given the
country’s low level of education.
12. In this way, politics in Honduras differs from that in Brazil. In Brazil, opposition
legislators ‘‘sell’’ their support for bills initiated by the executive in exchange for pork projects
and other resources (Ames 2001; Desposato 2001). They have this opportunity because the
president or governor rarely has a majority in the legislature, due to the fragmented party
system. The executive needs to ‘‘purchase’’ support in the legislature to pass bills, and deputies
can trade their votes for resources controlled by the executive branch. In Honduras, until 2002,
the president’s party always had a majority in Congress, and the president’s strong partisan
powers ensure backbencher loyalty. Under these circumstances, opposition deputies do not
have the opportunity to ‘‘sell’’ their votes to the executive in exchange for access to pork and
patronage, because the governing party has not needed their votes to pass legislation and does
not want to share resources.
13. Scott argues that an urban machine ‘‘dealt almost exclusively in particularistic, material
rewards to maintain and extend its control over its personnel. Although pork-barrel legislation
provided inducements for ethnic groups as a whole, the machine did most of its favors for
individuals and families’’ (1969, 1144). He also contends that patrons deliver many benefits to
clients during the enforcement stage of policy implementation, when the patron secures an
exception to a rule for the client.
14. This deputy is an outlier in many ways. He was a frequent initiator not only of locally
targeted bills but of all other types of bills as well. He also held leadership positions in Congress
and in his party. To make certain that this deputy and other Congress and party leaders were
not biasing the analysis, I ran all models excluding members of the congressional leadership,
and the results remained the same.
notes to pages 114 – 124 301
15. Since suplentes are not included in the analysis, a suplente elected for two consecutive
terms is not included in the count of reelected deputies. A suplente who became a propietario
in the next election was counted as reelected. A propietario reelected as a suplente was not
counted as reelected.
16. I also ran the analysis using other indicators: population with less than three years, and
less than six years, of schooling, and a composite indicator of the prosperity of a department
(based on the percentage of households with access to electricity, running water, and tele-
vision). The results did not change, and all indicators are strongly correlated.
17. The PINU and PDCH competed in the 1993 elections but did not win any munici-
palities. Small parties did not begin to win municipal elections until 1997.
18. There is one exception to the ‘‘wealthy’’ department–high population rule. The Bay
Islands are a small department in terms of population (DM = 1), but prosperous because of
tourism (only 11.6% of the population has no education). Despite its prosperity, in-migration to
that department has been low.
19. Hypothesis 1 predicts that politicians will perform more constituency service when they
are elected by open-list rather than closed-list rules. This is a routine expectation in the
electoral institution literature, but it cannot be tested here, as Honduras has always used
closed-list elections. The results thus begin with a test of hypothesis 2.
20. Most party and congressional leaders are elected from the two largest departments. In the
1994–97 Congress, the Congress president, three of four vice presidents, one of two secretaries,
and the faction leader of one of the major parties were elected from the two largest departments.
21. The only substantive difference between models 1 and 2 is that the department domi-
nance variable is not significant in model 1. However, theory does not predict a relationship
between district competitiveness and the incentive created by the size of DM in closed-list PR
electoral systems.
22. However, with a p value of .498, the sign for marginal list position has little meaning.
23. Simulations based on model 2 are almost identical to those for model 3.
24. However, overlapping confidence intervals between new and senior deputies make it
difficult to draw conclusions here.
25. All findings are substantively the same. The sign on average percentage of deputies
reelected in department becomes negative, but it never comes close to conventional signifi-
cance levels in any model.
26. There are signs that this support is eroding, as small parties won 12 of 128 seats in the 2001
elections, and for the first time even won seats outside the two largest departments. Even in the
2001 elections, however, the traditional parties won 80 percent of the vote in the congressional
elections and 92 percent in the presidential election.
27. This was most noticeable in the 1993 election, when PNH supporters were reported to
have stayed home to protest the lack of benefits from their party when it was in power, from
1990 to 1993. Both traditional parties should have learned from that election that while poor
clients may be reluctant to switch parties, they will not continue to vote for patrons who do not
deliver when they control executive-branch resources. This example of experiential learning,
even in long-established clientelistic political parties, illustrates how new ‘‘rules of the game’’
can be transmitted in informal institutions (Helmke and Levitsky, introduction to this vol-
ume). In this case, clients communicated rules to patrons.
302 notes to pages 125 – 138
1. One might object that grammatical rules do not govern, but simply describe, behavior.
Yet these rules, in fact, are as prescriptive as they are descriptive; the point of their formalization
is to police people’s usage, which is prone to stray from the rules.
2. I mean by accountability the ability and tendency of voters to turn out of office those
governments and politicians that, at the end of their term, are deemed by voters to have
performed below a threshold that the voters set. I mean by responsiveness a tendency for
governments and politicians to adopt policies preferred by (a majority of ) their constituents.
Democratic theorists often hold that accountability causes responsiveness, but responsiveness
can be achieved through other mechanisms as well.
3. On sociotropic and egocentric rules, see Ferejohn (1986); on socialization voting, Camp-
bell et al. (1964); on spatial voting, Downs (1957); on strategic voting, Cox (1997); on directional
voting, Rabinowitz and McDonald (1989); on ethnic voting, Chandra (2004); on clientelistic
voting, Stokes (2005).
4. Peronist party member, interview by Valeria Brusco, Marcelo Nazareno, and the author,
Córdoba, Argentina, January 2003.
5. Peronist party member, interview by Valeria Brusco, Marcelo Nazareno, and the author,
Córdoba, Argentina, January 2003.
6. See Gibson and Calvo (2000), Remmer and Wibbels (2000), UNDP (2002), Calvo and
Murillo (2004), Amaral and Stokes (2005), and Cleary and Stokes (forthcoming).
7. The surveys were carried out in December 2001 and January 2002. We instructed the
polling firm Consultores en Políticas Públicas, S.A., to conduct face-to-face interviews with
480 adults, aged eighteen or older, in each of the four regions. To select our sample, we used
multistage cluster sampling procedures, based on census tracks.
8. The models in table 6.2 include dummy variables for respondents who lived in all three
regions other than Mar del Plata. Negative signs on the Buenos Aires, Córdoba, and Misiones
dummies indicate that people in these provinces were less likely than others to offer ‘‘account-
ability’’ answers to the question.
9. Here and later I generated the simulations using the Clarify program (King et al. 2000;
Tomz et al. 2003). This program draws simulations of parameters of statistical models (in this
case, logit regressions) from their sampling distribution and then converts these simulated
parameters into expected values, such as expected probabilities of an answer to a survey
question, given hypothetical values of explanatory variables. For this simulation we held house-
hold income, educational level, quality of housing, age, and population size of the respon-
dents’ community at their sample means, and assumed a female Peronist supporter.
10. I did study the effect of the proportion of poor people in a municipality—measured as
either the proportion of houses that are substandard or the proportion of people in a munici-
pality with ‘‘unsatisfied basic needs’’—on the beliefs and expectations discussed in this chapter.
Unfortunately, these data come from the 1991 census and hence are not very reliable measures
of poverty rates in 2001–2, when we conducted the survey. Furthermore, poverty rates are a
very uncertain measure of the distribution of income: two communities with identical poverty
rates may have very different Gini indices, for example. With both caveats, the effect of poverty
notes to pages 143 – 154 303
rates was the reverse of what we might have expected: the higher the poverty rate in the
community, the more likely a person was (all else being equal) to offer ‘‘accountability’’
responses to various questions. Nor did poverty rates make regional effects disappear; indeed,
poverty rates, if anything, brought out these regional effects more strongly.
The author thanks Gretchen Helmke, Steven Levitsky, and Ignacio Marván Laborde for
their help with this chapter.
1. The PRI was created in 1929 by then-president Plutarco Elías Calles, and was placed on a
mass base by President Lázaro Cárdenas in the 1930s. For several decades, the party won
presidential elections by more than 70 percent of the national vote.
2. One obvious exception to this rule was Luis Donaldo Colosio, who was assassinated in a
campaign event after being named the PRI’s presidential candidate in the 1994 race.
3. The nonconsecutive reelection clause was changed several times before the final 1933
edict. In 1917, the representatives to the Constitutional Convention determined that there
would be no reelection for presidents and the term of office would last four years. In January
1927, General Álvaro Obregón changed the articles and allowed for the reelection of the
president after one term out of office, so revolutionary strongmen Álvaro Obregón and Plu-
tarco Elías Calles could alternate in the presidency. In January 1928, the presidential term was
lengthened to six years. In 1933, under the guidance of former president and then–jefe máximo
Elías Calles, the article was again reformed to prohibit consecutive reelection for congress-
men, local deputies, and municipal presidents and any reelection of presidents and governors
(including, for the first time, provisional and interim executives). With three constitutional
changes in fifteen years, it was entirely credible that the article could be changed once again at
the whim of a new president or strongman, so one cannot argue that the informal prerogatives
grew up around an entrenched formal rule. Cárdenas could have at least attempted to reverse
the nonreelection clause back to its 1927 phrasing to give him another chance to win the
presidency after a term out of office. The 1933 prohibition of reelection was not a rock-solid
formal institution until President Miguel Alemán (1946–52) challenged it by starting a whis-
pering campaign favoring his reelection (and a change in the Constitution), and failed (see
Nieto 1993). On Alemán’s failed attempt at reelection, see Paoli Bolio (1985).
4. The 1946 challenger had been the secretary of foreign relations in the then-current
administration.
5. In practice, only a few secretariats were considered important enough to field a possible
presidential candidate. These included the Treasury (Hacienda), Spending (Presupuesto),
Home Office (Gobernación), and, early on, the Labor ministry.
6. Some argue that de la Madrid and his aides simply did not take the CD leaders, Porfirio
Muñoz Ledo and Cuauhtémoc Cárdenas, seriously and, for this reason, never believed that
their break with the party would constitute such a serious threat to the electoral fortunes of the
PRI’s presidential candidate. Jorge de la Vega Domínguez, then-leader of the PRI, interview,
Mexico City, March 1996.
7. One should also point out that during the late 1990s, some changes in nomination
practices were not carried out through formal statutory reform—for example, gubernato-
304 notes to pages 155 – 164
rial nominations went from presidential imposition to open state primaries without any real
change in the formal party rules. The statutes were so ambiguous that the change could be
made without a formal rewriting.
8. Because of the particular electoral calendar in Mexico, with governors elected on a
staggered calendar, President Zedillo had not appointed many of the governors serving under
him in 1996. Few if any of the state executives were Zedillo allies, and they had no special
reason to protect his ability to place his own successor. Finally, two of the most independent
PRI governors, Manuel Bartlett of Puebla and Roberto Madrazo of Tabasco, attended the
statutory working group and helped instigate the rebellion that would remove the president’s
technocrats from the nomination race.
9. A large number of publications have described the changing nature of the regime’s elite
during the last two decades of its rule. The discussion centered on the growing importance of
technocratic, elite-educated leaders, who had never held elected posts and who rose to posi-
tions of power. The positions of bureaucrats and politicians were considered political careers
that held increasingly little promise (Camp 1985; Centeno 1994).
10. For more on Zedillo’s statements on the succession process, see Reforma (Mexico City),
June 10, 1998, and El Universal (Mexico City), October 13, 1998.
11. One should not exaggerate the fairness of the new nomination method. In Mexico, the
Federal Electoral Commission does not regulate party primaries, so each party must take
responsibility for managing its own nominations. This constitutes a problem when the de facto
leader of the party (in this case, President Zedillo) clearly has a favorite—and the resources to tip
the balance. The electoral method was designed to minimize the ability of any pre-candidate to
use fraud to win the contest (the winner had to capture a majority of the nation’s three hundred
electoral districts, so winning a million extra votes in a geographic region would not decide the
primary). Explicit rules to guarantee that nominees would receive equal treatment and re-
sources were, by and large, ignored, because the enforcement of these rules was largely
impossible. This is an important issue: to this day (2006), primaries are still not the only method
for choosing candidates, and there are no clear rules for the presidential nomination. This has
begun to provoke conflict within the PRI as groups look for leverage to choose the nominating
procedure for 2005. Formal rules still do not dictate all political outcomes within the PRI.
12. Closed primaries, consisting of only registered party members, could not be held
because there was no reliable membership list acceptable to the pre-candidates.
A related paper based on this research was originally published as ‘‘Insurance for Good
Losers and the Survival of Chile’s Concertación,’’ by John M. Carey and Peter M. Siavelis,
Latin American Politics and Society, vol. 47, no. 2 (Summer 2005), pp. 1–22.
1. This is true for the Chamber of Deputies, whose 120 members are elected from sixty two-
member districts, and for the elected members of the Senate, who are chosen, two each, from
nineteen districts.
2. The frequent result, as noted in many accounts of posttransition elections in Chile, is
that the second-place candidate in a district, in terms of preference votes, is not elected to
notes to pages 164 – 177 305
Congress. This happens when the first- and second-place candidates are on the same list, but
that list’s total does not double the total of the second-place list.
3. The Alianza por Chile has also been known as Democracia y Progreso and Unión por el
Progreso. The Union of the Center-Center (UCC) aligned formally with the Right for the 1993
election. Party key: PDC, Christian Democratic Party (Partido Demócrata Cristiano); PRSD,
Radical Social Democratic Party (Partido Radical Social Demócrata); PPD, Party for Democ-
racy (Partido Por La Democracia); PS, Socialist Party (Partido Socialista); RN, National Re-
newal (Renovación Nacional); UDI, Independent Democratic Union (Unión Demócrata
Independiente).
4. For a more comprehensive discussion of the intricacies of the Chilean party system and
the empirical conditions that bear on coalition formation, see Siavelis (2002b).
5. The mean shares of the two-coalition vote were, respectively, 59 and 41 percent, with a
standard deviation of 9 percent. Thus, the Concertación could double with a performance one
standard deviation above the mean, whereas the Right had to perform three standard devia-
tions above the mean.
6. The legislators wish to remain anonymous. The conversations took place in April 2003
with a former deputy from the Christian Democrats, and in December 2003 with an incum-
bent deputy from the UDI, the latter conducted by Adam Brinegar, who provided research
assistance for this project.
7. Concertación lists that successfully double, of course, produce no good losers, because
both their candidates are elected, so these lists are not included in the analysis.
8. www.elecciones.gov.cl/full/indexf.htm.
9. Including these Senate good losers should bias against supporting our model, as we are
beefing up the ‘‘least likelies’’ category with Senate candidates, who are more likely than
Chamber candidates to secure appointments.
10. ‘‘Most likelies’’ and ‘‘least likelies’’ were subject to an exhaustive search, beginning with
our appointment database, followed by an extensive search of the entire Government of Chile
website (including current and past legislators). We followed up with combined searches of
first name, last name, and political party, then first and last names, and then simply last names,
using the Google search engine (widely acknowledged as the most complete and accurate).
We examined every hit and followed up to ensure that we had identified the correct person,
paired with the correct postelection position. One could contend that the probable higher
visibility of ‘‘most likelies’’ makes them more likely to be found, skewing results in favor of our
hypothesis. To the extent that this was the case, however, our search efforts were asymmetric,
with greater effort devoted to the ‘‘least likelies,’’ for whom postelectoral career information was
generally more sparse. In the end, we are confident that the information we collected accu-
rately reflects the postelectoral fates of both groups.
11. UDI deputy, Congress of Chile, interview by Adam Brinegar, Valparaiso, December 9,
2003.
12. UDI deputy, Congress of Chile, interview by Adam Brinegar, Valparaiso, December 9,
2003.
13. Indeed, scrapping the two-member system has been on and off the Chilean agenda
since it was first imposed, over the objections of the parties that went on to form the Concerta-
306 notes to pages 178 – 204
ción. In the early 1990s, supporters of the former Pinochet regime defended this particular
formal institution against Concertación attacks. Over time, the Concertación adapted to
the two-member system, and its objections diminished. If the Chilean electoral landscape
changes, and Concertación dominance is shaken, the coalition’s discomfort with the two-
member system should revive.
1. Here we need to distinguish between an observer’s belief that the conduct in question
responds to a written rule and a shared belief among the relevant actors that their behavior is
required by an unwritten rule of some kind, which may be enforced if violated. As discussed in
the following paragraphs, the latter is precisely the sort of shared normative expectation that
defines an informal rule.
2. Some rules are enforced through a combination of negative sanctions for violations and
positive rewards for compliance. For a discussion of when a society might choose to reward
compliance rather than simply punish a breach, see Ellickson (1991, 124–26). For simplicity, in
notes to pages 205 – 210 307
this chapter I speak as if—as is usually the case—enforcement implies punishment rather than
reward, though it should be understood that they are merely opposite sides of the same coin.
3. In fact, it is the potential for enforcement behavior that makes informal institutions an
interesting concept. Without that, they add nothing to an explanation that could simply list the
various first-order incentives that cause actors to choose one course of conduct over another.
The corollary to this observation is that, just as we must define the organizational context in
which an informal rule operates, we must also identify the relevant enforcement instances that
are charged with ensuring that the informal institution is respected. It is these agents that must
hold some sort of normative attachment to the rule, in the relatively thin sense that they will
apply sanctions to deviant behavior in order to uphold the rule.
4. There are others who do explore the roots of the problem; see, e.g., Skolnick and Fyfe
(1993), Chevigny (1995), and Holston and Caldeira (1998).
5. I used a figure of thirty-six million inhabitants for the state of São Paulo, as reported on
the state government’s official website, and a figure of twelve million for the Buenos Aires
Metro area, from the Encyclopaedia Britannica’s list of largest metropolitan areas. The avail-
able figures for São Paulo suggest that the proportion of killings in the capital versus the rest of
the state roughly corresponds to the capital’s relative population.
6. This figure of 6.23 per 100,000 does not include an additional 85 annual victims of grupos
de extermínio, even though all accounts of these groups confirm a high participation of police
officers. See, e.g., interview with Costa Ferreira in Oiticica (2001); see also de Oliveira et al.
(2000).
7. The point is not that the conviction rate should be 100 percent. The other side of the coin
is the success of the institution in protecting the rights of those who are accused on the basis of
flimsy or inaccurate evidence. Thus a certain number of acquittals are to be expected and are,
indeed, a sign of health. But the rate of appropriate convictions should be higher where the
investigative, prosecutorial, and judicial functions are most effective.
8. Data on per capita police homicides are compiled from unofficial sources in Argentina
(Center for Legal and Social Studies [CELS] in Buenos Aires; Council against Police and
Institutional Repression [CORREPI] in Córdoba), and in São Paulo from the Ouvidoria da
Polícia, a state-run ombudsman organization. In Uruguay, data are from human rights reports,
especially the Peace and Justice Service (SERPAJ) and U.S. State Department, and my own
search of newspaper archives. In Salvador da Bahia, the data are from a study by the Arch-
diocese of Salvador and a number of other organizations. The conviction rates are from samples
of cases that I gathered and tracked through the courts. In Argentina, the points of departure for
the samples are lists of victims compiled by civil society organizations (primarily CORREPI)
that conduct daily newspaper reviews and have networks of informants in marginal areas. In
São Paulo, the starting point is a list compiled by the Ouvidoria da Polícia. In Uruguay, I
conducted my own search of newspaper archives, as well as reviewing all the available human
rights reports that dealt with police violence during the 1990s. In Salvador, I was unable to
piece together a representative sample of police homicide prosecutions. Instead, I interviewed
a number of representatives of NGOs, prosecutors, and lawyers, and reviewed newspaper
accounts, estimating a conviction rate about half that of São Paulo, based on my own qualita-
tive impression of the judicial response. Descriptive details of the performance of the courts in
Salvador that support this low evaluation are set forth at length in later sections of the chapter.
308 notes to pages 211 – 228
9. While the evidence suggests that the rule applies in Salvador as well, I have not included
that city in table 10.1, for two reasons. First, the evidence from Salvador is more qualitative and
not based on a representative sample, so simple percentages are misleading. Second, though it
includes Violent Victims, the rule seems broader in Salvador, as discussed in more detail in
the text.
10. Note that the conviction rates in non–Violent Victim cases are still quite low in Buenos
Aires and São Paulo. I have argued elsewhere (Brinks 2003a, 2003b) that these low conviction
rates are the result not of an informal rule but rather of other limitations of the legal system.
Due to space constraints, this chapter focuses on the Violent Victim cases in Buenos Aires and
São Paulo.
11. La Nación (Buenos Aires), October 4, 2000, p. 17.
12. Yulo Oiticica, Dossiê Grupos de Extermínio, unpublished legislative briefing material,
2000, on file with author.
13. Even after a separate judicial police force was authorized in the province of Buenos
Aires, investigators remained part of, and under the control and supervision of, the regular
police hierarchy. And in the federal system, the federal police continue to carry out all the
investigations and forensic tests required by the courts.
14. Case documents and interviews with Beatriz Sinisgalli of the Centro Santo Dias de
Direitos Humanos, São Paulo, 2001.
15. ‘‘Tras cinco años de concluída la tragedia, concluyó el juicio oral [Five years after the
tragedy, the oral trial concludes],’’ La Nación (Buenos Aires), December 8, 2004; and inter-
views with lawyers from CORREPI, 2000–2001.
16. ‘‘Policiais militares acusados de executar assaltante na Bahia camburtao [Military po-
lice accused of executing bank robber in police wagon],’’ A Tarde (Salvador), October 14, 1999.
17. ‘‘Mais que mil palavras [Worth a thousand words],’’ A Tarde (Salvador), October 13, 1999.
18. As of December 2005, extensive searches of the online archives of A Tarde do not reveal
any news items regarding a trial or conviction of these police officers.
19. ‘‘PM s matam jovem e ficam impunes [Military police kill youth with impunity],’’ A
Tarde (Salvador), December 31, 2000.
20. A Tarde (Salvador), April 16, 2000, p. 7.
21. Author’s review of case file, Case No. 1908-1998, vara 1a do júri, São Paulo.
22. José Antonio Carvalho, Cidade Repartida, Salvador, Gráfica do Sindicato dos Ban-
cários, Salvador, 2001, on file with author.
23. Valdemar Oliveira, director of CEDECA, interview, Salvador, Bahia, May 22, 2001;
CEDECA report, January 17, 2001, on file with author.
24. Ibid.
The author thanks Gretchen Helmke and Steven Levitsky for helpful comments, and
acknowledges research funding from the United States Agency for International Development
Grant 523-A-00-00-00030-00.
1. See, e.g., Aguilar Camín and Meyer (1993, 178–83) on consolidation of the PRI-state’s
Janus-faced policies toward dissenters, fully developed by the 1960s.
notes to pages 229 – 240 309
2. Between 1989 and 2000, the PRD staged 750 postelectoral conflicts, while the PAN
staged only 200 (Eisenstadt 2004, 141).
3. As PRD leader Javier González Garza (federal deputy and former national PRD director
of electoral affairs) put it: ‘‘The PRD is an optical illusion rather than a party,’’ and ‘‘every day,
there are a hundred acts in the country in the name of the party about which the national
directorate knows nothing.’’ Interview, Mexico City, August 25, 1996.
4. Any worthy governor followed this judicially inscrutable and politically expedient for-
mula, perfected, perhaps, by Puebla state’s ruthless Maximino Ávila Camacho (1937–41),
whose ‘‘intervention . . . in local elections assured that most of the municipalities would be
composed of allies sympathetic to the ‘political line’ of the governor,’’ prone to ‘‘dissolving
municipalities that remained outside his yoke, and naming, in their place, municipal council
governments’’ (Valencia Castrejón 1996, 106–7).
5. This database covers all mayoral races in Mexico’s thirty-one states between 1989 and
2000, based on coding of national and local print media. The numbers cited refer only to
municipal councils formed specifically to alleviate postelectoral conflicts (as opposed to those
formed to stem corruption, or because a mayor died, or for other reasons).
6. The federal documents used for this analysis were incomplete, accounting for only 103
municipal councils in fifteen of Mexico’s thirty-one states, between 1989 and 1996. Reasons for
creation of each municipal council were given in only 60 of these cases (Federal Secretary
of the Interior 1996). Percentages are based on an N of 60, which, as I argue elsewhere
(Eisenstadt 2004, 121), vastly underrepresents the actual number of occurrences.
7. An earlier ‘‘People’s Tribunal’’ in the state of Puebla was actually established as part of
the PAN’s Puebla 1983 postelectoral mobilization (Alvarez de Vicencio 1995, 142), but it did not
seem to have much transcendence.
8. Boylan quotes long-time Salinas confidant Manuel Camacho: ‘‘My impression is that
the president never stopped worrying about the possibility that [Cuauhtémoc] Cárdenas’
popularity might grow and that he might win in 1994 . . . This was true even after 1991 and
continued right until the very end’’ (2001a, 16).
9. Amador Rodríguez Lozano, former national PRI electoral affairs secretary and Baja Cali-
fornia senator, interview, Mexico City, August 12, 1996. Article 9 of the Yucatán state constitu-
tion calls for naming a temporary replacement, but only until a new election is convened.
10. Rodríguez Lozano, interview; Orlando Paredes, Yucatán gubernatorial candidate and
former Mérida mayoral candidate, interview, Mérida, Mexico, May 25, 2001.
11. An embittered Paredes denounced the agreement even in 2001, insisting that he had
never resigned. Interview, May 25, 2001.
12. The 1994 electoral process was deemed transparent by most observers, despite persistent
complaints, largely by the PRD, about subtle forms of manipulation of the electoral registry
and gerrymandering of congressional districts.
13. Andrés Manuel López Obrador, PRD Tabasco formal gubernatorial candidate, inter-
view, Villahermosa, Mexico, January 14, 1996.
14. Arturo Núñez Jiménez, former subsecretary of the Interior Secretariat of the Mexican
federal government, interview, Mexico City, August 28, 1996.
15. Jaime Cárdenas, former Federal Electoral Institute citizen counselor, interview, Mex-
ico City, July 19, 2000.
310 notes to pages 240 – 250
16. Sandra Fuentes-Berain, PRI international coordinator, with Carter Center delegation,
interview, Mexico City, June 14, 2000.
17. Fox was the 1991 Guanajuato gubernatorial candidate who contested his dubious loss
and negotiated an interim governorship for fellow Guanajuato PANista Carlos Medina Plas-
cencia (now a prominent senator). Interior secretary Santiago Creel, the most powerful cabi-
net member, was a clutch mediator in the attempted concertacesión with the PRD, Tabasco,
1994. And Senator Diego Fernández de Cevallos, the PAN’s interlocutor with Salinas during
most concertacesión negotiations of that era, was the PAN’s 1994 presidential candidate and
continues (as of 2006) to serve as one of the party’s most prominent senators. Federal controller
Francisco Barrio, probable victor of the Chihuahua 1986 governor’s race, tried to negotiate a
political solution to that fraud-riddled election but found no interlocutor in the presidential
administration of Miguel de la Madrid (1982–88).
18. Marco A. Zazueta Félix, PRI representative at the Federal Electoral Institute, interview,
Mexico City, July 18, 2000.
19. More than 10 percent of the PAN’s 321 legislative initiatives in the pre-1988 years were
electoral reform proposals, and it had advocated for the creation of an electoral court starting
in 1947 (Eisenstadt 204, 168).
20. The ruling was disseminated via the internet (www.trife.gob.mx).
21. Mauro Miguel Reyes Zapata, magistrate of the Electoral Court of the Judicial Power of
the Federation, interview, Mexico City, January 16, 2001.
22. For example, during the Tabasco controversy, Fox announced: ‘‘This is an issue for
Tabasco to decide and not for the Federal Executive, which has powers clearly delineated in
the constitution, which nowhere states that we should go around sticking our noses in to see
what happens in the states. Other presidents were accustomed to doing that, but it is not
happening now, nor will it’’ (Garduño 2001, 3).
23. José de Jesús Orozco Henríquez, magistrate of the Electoral Court of the Judicial Power
of the Federation, interview, Mexico City, January 18, 2001.
The author is indebted to Willem Assies and Raquel Yrigoyen for help with obtaining
documents, to Fernando García and René Kuppe for conveying information on the status of
relevant legislation, and to René Kuppe and Rachel Sieder for comments on a previous draft of
this chapter. She gratefully acknowledges funding support for field research from a 1997
Fulbright dissertation fellowship and two University of Tennessee Professional Development
Awards.
1. Where states officially have codified multiculturalism as a principle of the political
system and established regimes of legal pluralism, the category of competing is open to inter-
pretation. Sanctions imposed by culturally distinct indigenous authorities that may seem to
violate state law—such as corporal punishment—may be interpreted as not competing if one
evaluates them in the context of indigenous cultures, which are held to be equally valid
sources of law.
2. But in practice, as we shall see, those institutions that are more complementary than
notes to pages 250 – 283 311
substitutive tend also to be more accommodating than competing, since coordination and
cooperation are more likely to entail some accommodation. The converse is not true, however:
institutions closer to the substitutive end are not necessarily more competing, since they
typically have less contact with the state.
3. René Orellana Halkyer, interview, Santa Cruz, Bolivia, July 18, 1997; Isabelle Combes,
interview, Santa Cruz, Bolivia, July 18, 1997.
4. The accused is suspended in the air by his arms, which are tied behind his back.
5. Ronda violence must be evaluated in context. Starn (1999, 88–89) compares the approxi-
mately five thousand murders committed by police and the military in the south-central Andes
during the 1980s with only eleven ronda murders committed between 1976 and 1992, despite
the more than half a million ronderos active during that period.
6. Also, interviews with representatives of municipal authorities in Cuzco, Peru, July 2002.
7. All references in this section on juntas vecinales are to these Ministerio de Justicia y
Derechos Humanos publications (1997a, 1997b).
8. Costa Rica has produced a small amount of jurisprudence. However, whereas Costa
Rica’s Constitutional Court recognizes the private nature of indigenous authorities (Chacón
Castro 1999, 129), Colombia’s Court views them as public authorities.
conclusion
1. On variation in institutional strength in Latin America, see Levitsky and Murillo (2005).
2. For example, no recognizable (formal or informal) rule governed presidential mandates
in Argentina between 1928 and 2003. During that period, only two elected presidents com-
pleted their constitutional mandate, and both of them (Juan Perón and Carlos Menem)
modified the Constitution to grant themselves an additional term.
3. Other typologies of informal institutions in the developing world include those of Lauth
(2000), Hyden (2002), and Grzymala-Busse (2004).
4. Indeed, ‘‘clientelism may be the most prolific informal institution around the world’’
(Hyden 2002, 18).
5. According to Anna Grzymala-Busse (2004), patronage distribution and illicit party fi-
nance became particularly widespread in postsocialist countries that lacked robust partisan
competition.
6. See also Ledeneva (1998), Darden (2002), Collins (2003, 2004), and Gel’man (2004).
7. See also Griffiths (1986), Moore (1986), and Merry (1988).
8. According to K. Tsai, such informal norms emerge in ‘‘formal institutional environments
that fail to reflect the implicit demands of both political and economic actors’’ (2003, 3).
9. Interestingly, some Soviet-era norms survived the transition to a market economy (Clarke
1995; Ledeneva 1998; Sil 2001). As Rudra Sil (2001) has shown, for example, workers and
managers in many privatized Russian firms continued to adhere to preexisting ‘‘nonmarket
norms’’ during the 1990s in an effort to cope with new market institutions that they were unable
to change.
10. See, e.g., John Carey’s recent work (2003, 2004) on transparency in legislative voting.
11. As Siavelis suggests (this volume), Chile may be an exception in this regard.
312 notes to pages 283 – 289
12. Thus, whereas thirty-nine of forty-five sub-Saharan African countries were governed by
one-party regimes in 1989 (Olukoshi 1998, 9), five years later ‘‘single-party rule had been at least
formally abolished throughout almost the entire region’’ (Kirschke 2000, 386).
afterword
1. Which, for that matter, could also be South Chicago or the Bronx, which shows that
these problems do not refer only to ‘‘underdeveloped’’ countries. But the issue, to which I will
return, is the relative weight of the respective informal rules, and the degree to which, and the
number or type of occasions in which, they are effective.
2. On this subtle but important matter, see Bourdieu (1999).
3. Of course, I am not proposing this shifting actor perspective as an alternative to the study
of informal rules and institutions. I just believe that such a perspective would provide addi-
tional valuable information on this whole matter.
4. See the chapters by Siavelis, Mejía Acosta, Samuels, Taylor-Robinson, Carey and Sia-
velis, and Freidenberg and Levitsky.
5. This knowledge dates back to the authors of the Federalist Papers—and, for that matter,
even before them, to Montesquieu and Machiavelli, among others.
6. For the seminal work, see March and Simon (1958). A huge literature on various sides of
this matter followed.
7. See, e.g., Stokes’s chapter in this volume.
8. See the chapters by Desposato, Taylor-Robinson, Langston, Freidenberg and Levitsky,
Brinks, and Eisenstadt. For an apposite detailed study, see Hagopian (1996).
9. This remark points to a topic that has generated enormous attention and many polemics
in several disciplines and regions: the complex interactions between ‘‘imported’’ formal institu-
tions and developing countries’ institutions, rules, and practices. Of course, I cannot deal with
this matter here.
References
Abrúcio, Fernando. 1998. Os Barões da Federação. São Paulo, Brazil: Departamento de Ciên-
cia Política da Universidade de São Paulo–Editora Hucitec.
Acemoglu, Daron, and James Robinson. 2002. ‘‘The Political Economy of the Kuznets Curve.’’
Journal of Public Economics 84: 341–56.
————. 2005. Economic Origins of Dictatorship and Democracy. New York: Cambridge Univer-
sity Press.
Agüero, Felipe. 2003. ‘‘Chile: Unfinished Transition and Increased Political Competition.’’ In
Jorge I. Domínguez and Michael Shifter, eds., Constructing Democratic Governance in
Latin America. 2nd ed. Baltimore: Johns Hopkins University Press.
Aguilar Camín, Hector, and Lorenzo Meyer. 1993. In the Shadow of the Mexican Revolution—
Contemporary Mexican History, 1910–1989. Austin: University of Texas Press.
Akerlof, George A. 1970. ‘‘The Market for ‘Lemons’: Quality Uncertainty and the Market
Mechanism.’’ Quarterly Journal of Economics 84, no. 3: 488–500.
Alcántara Sáez, Manuel. 2004. Instituciones o máquinas ideológicas? Origen, programa y or-
ganización de los partidos latinoamericanos. Barcelona, Spain: Institut de Ciènces Poli-
tiques i Socials.
Alcántara Sáez, Manuel, and Flavia Freidenberg, eds. 2001. Partidos Políticos de América
Latina. Salamanca, Spain: Ediciones Universidad de Salamanca.
Alvarez de Vicencio, María Elena. 1995. Municipio y Democracia—Tesis y prácticas de go-
bierno del Partido Acción Nacional. Mexico City: Epessa.
Amaral, Samuel, and Susan C. Stokes. 2005. La democracia local en Argentina: Clientelismo,
capital social, e innovación en Argentina. Buenos Aires, Argentina: Universidad Tres de
Febrero.
Ames, Barry. 1987. Political Survival: Politicians and Public Policy in Latin America. California
Series on Social Choice and Political Economy 12. Berkeley: University of California Press.
————. 1995a. ‘‘Electoral Rules, Constituency Pressures, and Pork Barrel: Bases of Voting in the
Brazilian Congress.’’ Journal of Politics 57, no. 2: 324–43.
————. 1995b. ‘‘Electoral Strategy under Open-List Proportional Representation.’’ American
Journal of Political Science 39, no. 2: 406–33.
————. 2001. The Deadlock of Democracy in Brazil. Ann Arbor: University of Michigan Press.
Amorim Neto, Octavio. 1998. ‘‘Of President, Parties, and Ministers: Cabinet Formation and
Legislative Decision-making under Separation of Powers.’’ Ph.D. diss., Department of
Political Science, University of California, San Diego.
314 references
————. 2000. ‘‘Gabinetes presidenciais, ciclos eleitorais e disiplina legislative no Brasil.’’ Dados
43, no. 3: 479–519.
————. 2002. ‘‘Presidential Cabinets and Legislative Cohesion in Brazil.’’ In S. Morgenstern
and B. Nacif, eds., Legislative Politics in Latin America. Cambridge: Cambridge University
Press.
Amorim-Neto, Octavio, and Fabiano Santos. 2001. ‘‘The Executive Connection: Presidentially
Defined Factions and Party Discipline in Brazil.’’ Party Politics 7, no. 2: 213–34.
Araujo, Caridad. 1998. Gobernabilidad durante la crisis y políticas de ajuste. Quito, Ecuador:
Corporation of Development Studies (CORDES)–Konrad Adenauer Foundation.
Ardito Vega, Wilfredo. 2001. ‘‘Cambios culturales en la justica de paz.’’ Paper presented at Red
Latinoamericana de Antropología Jurídica-Sección Perú, Conferencia en Lima, Feb. 7.
Available at www.alertanet.org.
Assad, Carlos Martínez, and Alicia Ziccardi. 1988. Política y gestión municipal en México.
Mexico City: National Autonomous University of Mexico, Instituto de Investigaciones
Sociales.
Assies, Willem. 2001. ‘‘La oficialización de lo no oficial: (re)encuentro de dos mundos?’’
Alteridades 11, no. 21: 83–96.
————. 2003. ‘‘Indian Justice in the Andes: Re-rooting or Re-routing?’’ In Ton Salman and
Annalies Zoomers, eds., Imagining the Andes: Shifting Margins of a Marginal World?
Amsterdam: Center for Latin American Research and Documentation.
Austen-Smith, David, and Jeffrey Banks. 1990. ‘‘Stable Governments and the Allocation of
Policy Portfolios.’’ American Political Science Review 84, no. 3: 891–906.
Auyero, Javier. 2000a. ‘‘The Logic of Clientelism in Argentina: An Ethnographic Account.’’
Latin American Research Review 35, no. 3: 55–81.
————. 2000b. Poor People’s Politics: Peronist Survival Networks and the Legacy of Evita. Dur-
ham, NC: Duke University Press.
Axelrod, Robert. 1984. The Evolution of Cooperation. New York: Basic Books.
————. 1986. ‘‘An Evolutionary Approach to Norms.’’ American Political Science Review 80, no.
4: 1095–1112.
Bailey, John. 1988. Governing Mexico: The Statecraft of Crisis Management. New York: St.
Martin’s Press.
Baldez, Lisa, and John Carey. 1996. ‘‘The Chilean Budget Process.’’ Unpublished ms., paper
presented at the World Bank Conference on Policy Making in Latin America, University of
California, San Diego, May.
Banfield, Edward C., and James Q. Wilson. 1965. City Politics. Cambridge, MA: Harvard
University Press.
Barkin, David. 1990. Distorted Development—Mexico in the World Economy. Boulder, CO:
Westview Press.
Bauer, Raymond A., Alex Inkeles, and Clyde Kluckhorn. 1956. How the Soviet System Works:
Cultural, Psychological, and Social Themes. Cambridge, MA: Harvard University Press.
Berliner, Joseph S. 1957. Factory and Manager in the USSR. Cambridge, MA: Harvard Univer-
sity Press.
Bezerra, Marcos Otávio. 1999. Em Nome das ‘‘Bases’’: Política, Favor e Dependência Pessoal.
Rio de Janeiro, Brazil: Relume Dumará–Núcleo de Antropologia da Política.
references 315
Bill Chavez, Rebecca. 2004. The Rule of Law in Nascent Democracies: Judicial Politics in
Argentina. Palo Alto, CA: Stanford University Press.
Boeninger, Edgardo. 1989. ‘‘Gestión de gobierno y proceso de decisiones públicas.’’ Internal
memo circulated in Chilean presidential palace, Santiago.
Boix, Carles. 1999. ‘‘Setting the Rules of the Game: The Choice of Electoral Systems in
Advanced Democracies.’’ American Political Science Review 93, no. 3: 609–24.
————. 2003. Democracy and Redistribution. New York: Cambridge University Press.
Borocz, Joszef. 2000. ‘‘Informality Rules.’’ East European Politics and Societies 14, no. 2: 348–80.
Bourdieu, Pierre. 1999. ‘‘Rethinking the State: Genesis and Structure of the Bureaucratic
Field.’’ In George Steinmetz, ed., State/Culture: State Formation after the Cultural Turn.
Ithaca, NY: Cornell University Press.
Boussard, Caroline. 2000. ‘‘Democratic Consolidation: The Role of Informal Institutions:
Illustrations from Central America.’’ Paper presented at the twenty-second International
Congress of the Latin American Studies Association, Miami, Mar. 16–18.
Boylan, Delia. 1996. ‘‘Taxation and Transition: The Politics of the 1990 Chilean Tax Reform.’’
Latin American Research Review 31, no. 1: 7–31.
————. 2001a. ‘‘Democratization and Institutional Change in Mexico: The Logic of Partial
Insulation.’’ Comparative Political Studies 34, no. 1: 3–29.
————. 2001b. Diffusing Democracy: Central Bank Autonomy and the Transition from Authori-
tarian Rule. Ann Arbor: University of Michigan Press.
Bratton, Michael, and Nicolas van de Walle. 1994. ‘‘Neopatrimonial Regimes and Political
Transitions in Africa. World Politics 46, no. 4: 453–89.
Brazil Tribunal Superior Eleitoral. 2002a. ‘‘Divulagação de Prestação de Contas 2002.’’ Com-
puter files, Tribunal Superior Eleitoral, Brasília.
————. 2002b. ‘‘Resultados das Eleições de 2002.’’ Computer files, Tribunal Superior Eleitoral,
Brasília.
Brinks, Daniel. 2003a. ‘‘Courts, Norms, and Laws: Competing Rules and Legal Rights in
Salvador, São Paolo, and Buenos Aires.’’ Paper presented at the conference Informal In-
stitutions and Politics in Latin America, Kellogg Institute for International Studies, Univer-
sity of Notre Dame, IN, Apr. 24-25.
————. 2003b. ‘‘Informal Institutions and the Rule of Law: The Judicial Response to State
Killings in Buenos Aires and São Paolo in the 1990s.’’ Comparative Politics 36, no. 1: 1–19.
————. 2003c. ‘‘Legal Equality and Effectiveness in the Courts of Argentina, Uruguay and
Brazil.’’ Paper presented at the sixty-first annual meeting of the Midwest Political Science
Association, Chicago, Apr. 3–6.
Bruhn, Kathleen. 1977. Taking on Goliath: The Emergence of a New Left Party and the Struggle
for Democracy in Mexico. University Park: Pennsylvania State University Press.
Brusco, Valeria, Marcelo Nazereno, and Susan C. Stokes. 2004. ‘‘Vote Buying in Argentina.’’
Latin American Research Review 39, no. 2: 66–88.
Burbano de Lara, Felipe, and Michel Rowland. 1998. Pugna de Poderes: Presidencialismo y
Partidos en el Ecuador: 1979–1997. Quito, Ecuador: Corporation of Development Studies
(CORDES)–Konrad Adenauer Foundation and Spanish Cooperation.
Butler, Edgar W., and Jorge Bustamante, eds. 1991. Sucesión Presidencial: The 1988 Presidential
Election. Boulder, CO: Westview Press.
316 references
Cabedo Mallol, Vincente José. 1998. ‘‘El Pluralismo Jurídico en Colombia, Peru, Bolivia y
Ecuador: Derecho Consuetudinario y Jurisdicción Indígena.’’ Cuadernos constitucionales
(Valencia, Spain), no. 25 (autumn). Available at www.geocities.com/alertanet.
Cain, Bruce E., John Ferejohn, and Morris Fiorina. 1987. The Personal Vote: Constituency
Service and Electoral Independence. Cambridge, MA: Harvard University Press.
Calvert, Randall. 1995. ‘‘The Rational Choice Theory of Social Institutions: Cooperation,
Coordination, and Communication.’’ In Jeffrey S. Banks and Eric A. Hanushek, eds.,
Modern Political Economy. New York: Cambridge University Press.
Calvo, Ernesto, and Victoria Murillo. 2004. ‘‘Who Delivers? Partisan Clients in the Argentine
Electoral Market.’’ American Journal of Political Science 48, no. 4: 742–57.
Camp, Roderic A. 1980. Mexico’s Leaders, Their Education and Recruitment. Tucson: Univer-
sity of Arizona Press.
————. 1985. ‘‘The Political Technocrat in Mexico and the Survival of the Political System.’’
Latin American Research Review 20, no. 1: 97–118.
Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes. 1964. The
American Voter. New York: John Wiley and Sons.
Carey, John M. 1996. Term Limits and Legislative Representation. New York: Cambridge
University Press.
————. 2000. ‘‘Parchment, Equilibria, and Institutions.’’ Comparative Political Studies 33, nos.
6–7: 735–61.
————. 2002. ‘‘Parties and Coalitions in Chile in the 1990s.’’ In Scott Morgenstern and Benito
Nacif, eds., Legislative Politics in Latin America. Cambridge: Cambridge University Press.
————. 2003. ‘‘Transparency versus Collective Action: The Fujimori Legacy and the Peruvian
Congress.’’ Comparative Political Studies 36, no. 9: 983–1006.
————. 2004. ‘‘Visible Votes: Recorded Voting and Legislative Accountability in Latin Amer-
ica.’’ Unpublished ms., Dartmouth College, Hanover, NH.
Carey, John M., and Matthew Soberg Shugart. 1995. ‘‘Incentives to Cultivate a Personal Vote:
A Rank Ordering of Electoral Formulas.’’ Electoral Studies 14, no. 4: 417–39.
Carey, John M., and Peter Siavelis. 2003. ‘‘El ‘seguro’ para los subcampeones y la sobrevivencia
la la Concertación.’’ Estudios Políticos 90 (fall): 5–27.
————. 2005. ‘‘Insurance for Good Losers and the Survival of Chile’s Concertación.’’ Latin
American Politics and Society 47, no. 2: 1–22.
Castañeda, Jorge G. 1999. La herencia: Arqueología de la sucesión presidencial en México.
Mexico City: Extra Alfaguara.
Castillo, Pilar del, and Daniel Zovatto, eds. 1998. La Financiación de la Política en Ibero-
américa. San José, Costa Rica: Instituto Interamericano de Derechos Humanos.
Centeno, Miguel. 1994. Democracy within Reason: Technocratic Revolution in Mexico. Univer-
sity Park: Pennsylvania State University Press.
Centro de Estudios Públicos. 2003. ‘‘Puntos Salientes: Encuesta nacional de opinión pública.’’
Santiago, Chile: Centro de Estudios Públicos. Available at www.cepchile.cl.
Chacón Castro, Rubén. 1999. ‘‘Comentario a la presentación de Rodolfo Piza E.: Análisis de
jurisprudencia constitucional emergente en materia indígena en el caso de Costa Rica:
poder estatal y pueblos indígenas.’’ In Memoria: II Seminario Internacional sobre Admin-
references 317
istración de Justicia y Pueblos Indígenas. San José, Costa Rica: Instituto Interamericano de
Derechos Humanos.
Chambers, Ian. 1997. Untitled comments. In Magdalena Gómez, ed., Derecho Indígena.
Mexico City: Instituto Nacional Indigenista.
Chandra, Kanchan. 2004. Why Ethnic Parties Succeed: Patronage and Ethnic Headcounts in
India. Cambridge: Cambridge University Press.
Chevigny, Paul. 1995. Edge of the Knife: Police Violence in the Americas. New York: New Press.
Chull Shin, Doh. 1994. ‘‘On the Third Wave of Democratization: A Synthesis and Evaluation
of Recent Theory and Research.’’ World Politics 47, no. 1: 135-70.
Clarke, Simon, ed. 1995. Management and Industry in Russia: Formal and Informal Relations
in the Period of Transition. Aldershot, England: Edward Elgar.
Cleary, Matthew, and Susan C. Stokes. Forthcoming. Democracy and the Culture of Skepti-
cism: Political Trust in Argentina and Mexico. New York: Russell Sage Foundation.
Colignon, Richard A., and Chikako Usui. 2003. Amadukari: The Hidden Fabric of Japan’s
Economy. Ithaca, NY: Cornell University Press.
Collier, Jane. 1998. ‘‘The Peaceful Resolution of a Cortacabezas Case in the Toztzil Maya
Community of Zinacantan, Chiapas, México.’’ América Indígena 58, nos. 1–2: 203–20.
Collins, Kathleen. 2002a. ‘‘Clans, Pacts and Politics in Central Asia.’’ Journal of Democracy 13,
no. 3: 137–52.
————. 2002b. ‘‘Clans, Pacts, and Regime Transitions in Central Asia.’’ Paper presented at the
conference Informal Institutions and Politics in the Developing World, Harvard Univer-
sity, Cambridge, MA, Apr. 5–6.
————. 2003. ‘‘The Political Role of Clans in Central Asia.’’ Comparative Politics 35, no. 2: 171–90.
————. 2004. ‘‘The Logic of Clan Politics: Evidence from the Central Asian Trajectories.’’
World Politics 56, no. 2: 224–61.
Conaghan, Catherine. 1995. ‘‘Politicians against Parties: Discord and Disconnection in Ec-
uador’s Party System.’’ In Scott Mainwaring and Timothy R. Scully, eds., Building Demo-
cratic Institutions: Party Systems in Latin America. Stanford, CA: Stanford University Press.
Congreso Nacional del Ecuador. 1995. Impeachment Proceedings of Vice-President Alberto
Dahik. Quito, Ecuador: Congreso Nacional del Ecuador.
Cordova, Arnaldo. 1992. La formación del poder político en México. Mexico City: Era Pub-
lishers.
Cornelius, Wayne A. 1996. Mexican Politics in Transition: The Breakdown of a One-Party-
Dominant Regime. Monograph Series, 41. San Diego: University of California, Center for
U.S.-Mexican Studies.
Cox, Gary. 1987. The Efficient Secret. New York: Cambridge University Press.
————. 1997. Making Votes Count. Cambridge: Cambridge University Press.
Cox, Gary W., and Scott Morgenstern. 2002a. ‘‘Epilogue: Latin America’s Reactive Assemblies
and Proactive Presidents.’’ In Scott Morgenstern and Benito Nacif, eds., Legislative Politics
in Latin America. Cambridge: Cambridge University Press.
————. 2002b. ‘‘Reactive Assemblies and Proactive Presidents: A Typology of Latin American
Presidents and Legislatures.’’ In Scott Morgenstern and Benito Nacif, eds., Legislative
Politics in Latin America. Cambridge: Cambridge University Press.
318 references
Cox, Gary, and Michael Thies. 1998. ‘‘The Cost of Intra-Party Competition: The Single, Non-
transferable Vote and Money Politics in Japan.’’ Comparative Political Studies 31 (June):
267–91.
————. 2000. ‘‘How Much Does Money Matter? ‘Buying’ Votes in Japan, 1967-1990.’’ Compara-
tive Political Studies 33, no. 1: 37–57.
Craig, Ann L., and Wayne A. Cornelius. 1995. ‘‘Houses Divided: Parties and Political Reform
in Mexico.’’ In Scott Mainwaring and Timothy R. Scully, eds., Building Democratic In-
stitutions: Party Systems in Latin America. Stanford, CA: Stanford University Press.
Crain, W. Mark. 1977. ‘‘On the Structure and Stability of Political Markets.’’ Journal of Political
Economy 85 (Aug.): 829–42.
Crawford, Sue E. S., and Elinor Ostrom. 1995. ‘‘A Grammar of Institutions.’’ American Political
Science Review 89, no. 3: 582–600.
Darden, Keith. 2002. ‘‘Graft and Governance: Corruption as an Informal Mechanism of State
Control.’’ Paper presented at the conference Informal Institutions and Politics in the De-
veloping World, Weatherhead Center for International Affairs, Harvard University, Cam-
bridge, MA, Apr. 5–6.
Degregori, Carlos Iván, and María Ponce Mariños. 2000. ‘‘Movimientos sociales y estado: El
caso de las rondas campesinas de Cajamarca y Piura.’’ In Carlos Iván Degregori, ed., No
Hay País Más Diverso: Compendio de antrolopología peruana. Lima: Pontificia Univer-
sidad Catolíca del Peru, Instituto de Estudios Peruanos.
Deheza, Grace Ivana. 1997. ‘‘Gobiernos de coalición en el sistema presidencial: América del
Sur.’’ Ph.D. diss., Political Science Department, European University Institute, Florence,
Italy.
De la Torre, Augusto, Roberto García-Saltos, and Yira Mascaró. 2001. Banking, Currency, and
Debt Meltdown: Ecuador Crisis in the Late 1990s. World Bank working paper. Washington,
DC: World Bank.
Della Porta, Donatella, and Alberto Vannucci. 1999. Corrupt Exchanges: Actors, Resources,
and Mechanisms of Political Corruption. New York: Aldine de Gruyter.
De Luca, Miguel, Mark P. Jones, and María Inés Tula. 2002. ‘‘Back Rooms or Ballot Boxes?
Candidate Nomination in Argentina.’’ Comparative Political Studies 35, no. 4: 413–36.
de Oliveira, Nelson, Lutz Mulert Sousa Ribeiro, and Jose Carlos Zanetti, eds. 2000. A Outra
Face da Moeda. Salvador, Bahia, Brazil: CommisSão de Justiça e Paz da Arquidiocese de
Salvador.
Desposato, Scott W. 2001. ‘‘Institutional Theories, Societal Realities, and Party Politics in
Brazil.’’ Ph.D. diss., University of California, Los Angeles.
————. 2004. ‘‘The Impact of Federalism on National Political Parties in Brazil.’’ Legislative
Studies Quarterly 29: 259–85.
Dia, Mamadou. 1996. Africa’s Management in the 1990s: Reconciling Indigenous and Trans-
planted Institutions. Washington, DC: World Bank.
Domínguez, Jorge I., and Michael Shifter, eds. 2003. Constructing Democratic Governance in
Latin America. 2nd ed. Baltimore: Johns Hopkins University Press.
Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper and Row.
Duverger, Maurice. 1954. Political Parties: Their Organization and Activity in the Modern
State. New York: John Wiley and Sons.
references 319
Easter, Gerald. 2000. Reconstructing the State: Personal Networks and Elite Identity in Soviet
Russia. New York: Cambridge University Press.
Economist Intelligence Unit. 2000. Country Profile: Ecuador. London: Economist Intelli-
gence Unit.
Eisenstadt, Todd A. 1999. ‘‘Electoral Federalism or Abdication of Presidential Authority? Gu-
bernatorial Elections in Tabasco.’’ In Wayne Cornelius, Todd Eisenstadt, and Jane Hind-
ley, eds., Subnational Politics and Democratization in Mexico. La Jolla, CA: Center for
U.S.-Mexican Studies.
————. 2003. ‘‘Thinking Outside the (Ballot) Box: Mexico’s Informal Electoral Institutions.’’
Latin American Politics and Society 45, no. 1: 25–54.
————. 2004. Courting Democracy in Mexico. Cambridge: Cambridge University Press.
Ellickson, Robert C. 1991. Order without Law: How Neighbors Settle Disputes. Cambridge,
MA: Harvard University Press.
Elster, Jon. 1989. The Cement of Society: A Survey of Social Order. New York: Cambridge
University Press.
Epoca, La (Santiago). 1993a. ‘‘RN anunció definitivamente no dará más concesiones a la
UDI.’’ Mar. 27.
————. 1993b. ‘‘La DC quiere que Eduardo Frei sea el árbitro del pacto parlamentario.’’ June 22.
Epstein, Lee, and Jack Knight. 1998. The Choices Judges Make. Washington, DC: Congressio-
nal Quarterly Press.
Erdmann, Gero. 2004. ‘‘Party Research: Western European Bias and the ‘African Labyrinth.’ ’’
Democratization 11, no. 3: 63–87.
Estévez, Federico, Beatriz Magaloni, and Alberto Diaz-Cayeros. 2003. A Portfolio Diversifica-
tion Model of Policy Choice. Paper presented at the conference Clientelism in Latin
America: Theoretical and Comparative Perspectives, Stanford University, CA, May 17–18.
Euraque, Darío E. 2000. ‘‘Los políticos hondureños y la Costa Norte (1876-1950): narrativa e
interpretación.’’ Revista Política de Honduras 2: 113–56.
Farrell, Henry, and Adrienne Héritier. 2002. ‘‘Formal and Informal Institutions under Codeci-
sion: Continuous Constitution Building in Europe.’’ European Integration Online Papers
6, no. 3. Available at http://eiop.or.at/erpa/OAI/eiopxx/EIoP-P0076.html.
Fearon, James. 1999. ‘‘Electoral Accountability and the Control of Politicians: Selecting Good
Types versus Sanctioning Poor Performance.’’ In Adam Przeworski, Susan C. Stokes, and
Bernard Manin, eds., Democracy, Accountability, and Representation. Cambridge: Cam-
bridge University Press.
Federal Secretary of the Interior, Mexico. 1996. ‘‘Relación de Concejos Municipales Instala-
dos en el País Durante los Ultimos Tres Periodos de los Ayuntamientos.’’ Internal ms.,
Secretaría de Gobernación, Mexico City.
Fenno, Richard F., Jr. 1978. Homestyle. Boston: Little Brown.
Ferejohn, John. 1986. ‘‘Incumbent Performance and Electoral Control.’’ Public Choice 50,
nos. 1–3: 5–25.
————. 1999. ‘‘Accountability and Authority: Toward a Theory of Political Accountability.’’ In
Adam Przeworski, Susan C. Stokes, and Bernard Manin, eds., Democracy, Accountability,
and Representation. Cambridge: Cambridge University Press.
Ferreira Rubio, Delia, ed. 1997. Financiamiento de Partidos Políticos. Buenos Aires, Argentina:
320 references
tions in Taiwan’s Political Process.’’ Paper presented at the annual meeting of the American
Political Science Association, San Francisco, Aug. 30–Sept. 2.
Gómez, Leopoldo. 1991. ‘‘Elections, Legitimacy, and Political Change in Mexico, 1977–1988.’’
Ph.D. diss., Georgetown University, Washington, DC.
Gómez, Magdalena. 2000. ‘‘Derecho Indígena y constitucionalidad.’’ Paper presented at the
twelfth International Congress of the Comisión de Derecho Consuetudinario y Pluralismo
Legal, Arica, Chile, Mar. 13–17. Available at www.geocities.com/alertanet/index.html.
González Casanova, Pablo. 1965. La democracia en México. Mexico City: Era Publishers.
————. 1981. El Estado y los partidos políticos en México. Mexico City: Era Publishers.
Graeff, E. 2000. ‘‘The Flight of the Beetle: Party Politics and the Decision-Making Process in
the Cardoso Government.’’ Paper presented at the fifth Congress of the Brazilian Studies
Association, Recife, Brazil, June.
Graham, Richard. 1990. Patronage and Politics in Nineteenth-Century Brazil. Stanford, CA:
Stanford University Press.
Granados Chapa, Miguel Angel. 2001. ‘‘Jalisco en el Trife.’’ Reforma (Mexico City), Feb. 20, p.
15A.
Greif, Avner, and David D. Laitin. 2004. ‘‘A Theory of Endogenous Institutional Change.’’
American Political Science Review 98, no. 4: 633–53.
Griffiths, John. 1986. ‘‘What Is Legal Pluralism?’’ Journal of Legal Pluralism and Unofficial
Law 24: 1–55.
Grindle, Merilee, and Francisco E. Thoumi. 1993. ‘‘Muddling toward Adjustment: The Politi-
cal Economy of Economic Policy Change in Ecuador.’’ In R. H. Bates and A. O. Krueger,
eds., Political and Economic Interactions in Economic Policy Reform: Evidence from Eight
Countries. Oxford: Blackwell.
Grzymala-Busse, Anna. 2004. ‘‘Informal Institutions and the Post-Communist State.’’ Un-
published ms., Department of Political Science, Yale University, New Haven, CT.
Grzymala-Busse, Anna, and Pauline Jones Luong. 2002. ‘‘Reconceptualizing the State: Les-
sons from Post-Communism.’’ Politics and Society 30, no. 4: 529–54.
Hagopian, Frances. 1993. ‘‘After Regime Change: Authoritarian Legacies, Political Represen-
tation, and the Democratic Future of South America.’’ World Politics 45, no. 3: 464–500.
————. 1996. Traditional Politics and Regime Change in Brazil. New York: Cambridge Univer-
sity Press.
————. 1998. ‘‘Democracy and Political Representation in Latin America in the 1990s: Pause,
Reorganization, or Decline?’’ In Felipe Agüero and Jeffrey Stark, eds., Fault Lines of
Democracy in Post-Transition Latin America. Miami: North-South Center Press.
Hale, Henry E. 2005. ‘‘Why Not Parties? Electoral Markets, Party Substitutes, and Stalled
Democratization in Russia.’’ Comparative Politics 37, no. 2: 147-66.
Hamilton-Hart, Natasha. 2000. ‘‘The Singapore State Revisited.’’ Pacific Review 13, no. 2:
195–216.
Handy, Jim. 2004. ‘‘Chicken Thieves, Witches, and Judges: Vigilante Justice and Customary
Law in Guatemala.’’ Journal of Latin American Studies 36, no. 3: 533–61.
Hart, H. L. A. 1961. The Concept of Law. New York: Oxford University Press.
Hartlyn, Jonathan. 1998. The Struggle for Democratic Politics in the Dominican Republic.
Chapel Hill: University of North Carolina Press.
322 references
Heath, Roseanna Michelle, and Michelle M. Taylor-Robinson. 2003. ‘‘All Dressed up with No
Place to Go? Political Ambition in Unitary Political Systems with Unicameral Legisla-
tures.’’ Paper presented at the annual meeting of the American Political Science Associa-
tion, Philadelphia, Aug. 28–31.
Helmke, Gretchen. 2002. ‘‘The Logic of Strategic Defection: Court-Executive Relations in
Argentina under Dictatorship and Democracy.’’ American Political Science Review 96, no.
2: 291–304.
————. 2004. Courts under Constraints: Judges, Generals, and Presidents in Argentina. New
York: Cambridge University Press.
Helmke, Gretchen, and Steven Levitsky. 2003. ‘‘Informal Institutions and Comparative Poli-
tics: A Research Agenda.’’ Paper presented at the conference Informal Institutions in Latin
America, University of Notre Dame, IN, Apr. 23–24.
————. 2004. ‘‘Informal Institutions and Comparative Politics: A Research Agenda.’’ Perspectives
on Politics 2, no. 4: 725–40.
Hey, Jeanne A. K., and Thomas Klak. 1999. ‘‘From Protectionism towards Neoliberalism:
Ecuador across Four Administrations (1981–1996).’’ Studies in Comparative International
Development 34, no. 3: 66–97.
Hillman, Richard. 1994. Democracy for the Privileged: Crisis and Transition in Venezuela.
Boulder, CO: Lynne Reinner.
Holston, James, and Teresa P. R. Caldeira. 1998. ‘‘Democracy, Law, and Violence: Disjunc-
tions of Brazilian Citizenship.’’ In Felipe Agüero and Jeffrey Stark, eds., Fault Lines of
Democracy in Post-Transition Latin America. Coral Gables, FL: North-South Center Press.
Honaker, James, Anne Joseph, Gary King, Kenneth Scheve, and Naunihal Singh. 2001. Ame-
lia: A Program for Missing Data. Windows version. Harvard University. Available at
http://gking.harvard.edu.
Hooker, M. B. 1975. Legal Pluralism: An Introduction to Colonial and Neo-Colonial Laws.
Oxford: Clarendon Press.
Huntington, Samuel P. 1968. Political Order in Changing Societies. New Haven, CT: Yale
University Press.
Hurtado, Osvaldo. 1990. Política Democrática: los últimos veinte y cinco años. Quito, Ecuador:
Fundación Ecuatoriana de Estudios Sociales y Corporación Editora Nacional.
Hyden, Goran. 1980. Beyond Ujamaa in Tanzania: Underdevelopment and an Uncaptured
Peasantry. London: Heinemann.
————. 2002. ‘‘Why Africa Finds It So Hard to Develop.’’ Paper presented at the conference
Culture, Democracy, and Development: Cultural and Political Foundations of Socio-
Economic Development in Africa and Asia, Monte Verita, Ascona, Switzerland, Oct. 6–11.
Jackson, Jean. 2002. ‘‘Caught in the Crossfire: Colombia’s Indigenous Peoples during the
1990s.’’ In David Maybury-Lewis, ed., The Politics of Ethnicity: Indigenous Peoples in Latin
American States. Cambridge, MA: Harvard University Press.
Johannes, John R. 1984. To Serve the People: Congress and Constituency Service. Lincoln:
University of Nebraska Press.
Johnson, Kenneth. 1971. Mexican Democracy: A Critical View. Boston: Allyn and Bacon.
Jones, Mark. 1995. Electoral Laws and the Survival of Presidential Democracies. Notre Dame,
IN: University of Notre Dame Press.
references 323
Katz, Richard S., and Peter Mair, eds. 1994. How Parties Organize: Change and Adaptation in
Party Organizations in Western Democracies. London: Sage.
————. 1995. ‘‘Changing Models of Party Democracy: The Emergence of the Cartel Party.’’
Party Politics 1, no. 1: 5–28.
Katzenstein, Peter J. 1996. Cultural Norms and National Security: Police and Military in
Postwar Japan. Ithaca, NY: Cornell University Press.
Keech, William R. 1995. Economic Politics: The Costs of Democracy. New York: Cambridge
University Press.
Kenney, Charles. 2004. Fujimori’s Coup and the Breakdown of Democracy in Latin America.
Notre Dame, IN: University of Notre Dame Press.
King, Gary, Michael Tomz, and Jason Wittenberg. 2000. ‘‘Making the Most of Statistical
Analyses: Improving Interpretation and Presentation.’’ American Journal of Political Sci-
ence 44, no. 2: 347–55.
King, Gary, James Honaker, Anne Joseph, and Kenneth Scheve. 2001. ‘‘Analyzing Incomplete
Political Science Data: An Alternative Algorithm for Multiple Imputation.’’ American
Political Science Review 95, no. 1: 49–69.
Kirchheimer, Otto. 1966. ‘‘The Transformation of Western European Party Systems.’’ In Jo-
seph LaPalombara and Myron Weiner, eds., Political Parties and Political Development.
Princeton, NJ: Princeton University Press.
Kirschke, Linda. 2000. ‘‘Informal Repression, Zero-Sum Politics and Late Third Wave Transi-
tions.’’ Journal of Modern African Studies 38, no. 3: 383-405.
Kitschelt, Herbert. 1994. The Transformation of European Social Democracy. Cambridge:
Cambridge University Press.
————. 2000. ‘‘Linkages between Citizens and Politicians in Democratic Politics.’’ Comparative
Political Studies 33, nos. 6–7: 845–79.
Klesner, Joseph L. 1988. ‘‘Electoral Reform in an Authoritarian Regime: The Case of Mexico.’’
Ph.D. diss., Massachusetts Institute of Technology, Cambridge.
Knight, Alan. 1990. ‘‘Mexico, c. 1930–1946.’’ In Leslie Bethell, ed., Cambridge History of Latin
America, vol. 7. Cambridge: Cambridge University Press.
————. 1992. ‘‘Mexico’s Elite Settlement: Conjuncture and Consequences.’’ In John Higley and
Richard Gunther, eds., Elites and Democratic Consolidation in Latin America and South-
ern Europe. Cambridge: Cambridge University Press.
Knight, Jack. 1992. Institutions and Social Conflict. New York: Cambridge University Press.
Kroszner, Randall, and Thomas Stratmann. 1998. ‘‘Interest-Group Competition and the Orga-
nization of Congress: Theory and Evidence from Financial Services’ Political Action
Committees.’’ American Economic Review 88 (Dec.): 1163–87.
Kymlicka, Will. 1995. Multicultural Citizenship. Oxford: Oxford University Press.
Lancaster, Thomas D. 1986. ‘‘Electoral Structures and Pork Barrel Politics.’’ International
Political Science Review 7, no. 1: 67–81.
Landau, David. 2001. ‘‘Policy or Pork? Presidential Strategy Choice and the Legislature in
Ecuador.’’ B.A. thesis, Faculty of Arts and Sciences, Harvard University, Cambridge, MA.
Langston, Joy. 2001.‘‘ Why Rules Matter: Changes in Candidate Selection in Mexico’s PRI,
1988–2000.’’ Journal of Latin American Studies 33, no. 3: 485-512.
————. 2003. ‘‘The Formal Bases of Informal Power: Mexico’s PRI.’’ Paper presented at the
324 references
conference Informal Institutions and Politics in Latin America, Kellogg Institute for Inter-
national Studies, University of Notre Dame, IN, Apr. 24–25.
Latin America Advisor. 2002. ‘‘Chilean Lawmakers Decry Leader’s Statement on Ruling Coali-
tion.’’ Washington, DC: Inter-American Dialogue. Dec. 11.
Lau, Rubén. 1989. Cuadernos del Norte—Las Elecciones en Chihuahua (1983-1988). Chihua-
hua, Mexico: Centro de Estudios Libres Ignacio Rodríguez Terrazas, A.C.
Lauth, Hans-Joachim. 2000. ‘‘Informal Institutions and Democracy.’’ Democratization 7, no. 4:
21–50.
Laver, Michael, and Kenneth Shepsle. 1990. ‘‘Coalitions and Cabinet Government.’’ Ameri-
can Political Science Review 84, no. 3: 873–90.
Lawson, Chappell H. 2002. Building the Fourth Estate: Democratization and the Rise of a Free
Press in Mexico. Berkeley: University of California Press.
Ledeneva, Alena A. 1998. Russia’s Economy of Favors: Blat, Networking and Informal Ex-
change. Cambridge: Cambridge University Press.
Lemarchand, René. 1981. ‘‘Comparative Political Clientelism: Structure, Process and Optic.’’
In S. N. Eisenstadt and René Lemarchand, eds., Political Clientelism, Patronage and
Development. Beverly Hills, CA: Sage.
Lemos-Nelson, Ana Tereza. 2001. ‘‘Judiciary Police Accountability for Gross Human Rights
Violations: The Case of Bahia, Brazil.’’ Ph.D. diss., University of Notre Dame, Notre
Dame, IN.
Lerner de Sheinbaum, Bertha, and Susan Ralsky. 1976. El poder de los presidentes. Mex-
ico City: National Autonomous University of Mexico, Instituto Mexicano de Estudios
Políticos.
Levi, Margaret. 1997. Consent, Dissent, and Patriotism. New York: Cambridge University
Press.
Levitsky, Steven. 2003a. ‘‘Chaos and Renovation: Institutional Weakness and the Transforma-
tion of Argentine Peronism, 1983–2002.’’ Paper presented at the conference Rethinking
Dual Transitions: Argentine Politics in the 1990s in Comparative Perspective, Weatherhead
Center for International Affairs, Harvard University, Cambridge, MA, Mar. 20–22.
————. 2003b. Transforming Labor-Based Parties in Latin America: Argentine Peronism in Com-
parative Perspective. New York: Cambridge University Press.
Levitsky, Steven, and M. Victoria Murillo. 2005. ‘‘Theorizing about Weak Institutions: Lessons
from the Argentine Case.’’ In Steven Levitsky and M. Victoria Murillo, eds., Argentine
Democracy: The Politics of Institutional Weakness. University Park: Pennsylvania State
University Press.
Levitsky, Steven, and Lucan A. Way. 2002. ‘‘The Rise of Competitive Authoritarianism.’’ Jour-
nal of Democracy 13, no. 2: 51–65.
Lewin, Linda. 1987. Politics and Parentela in Paraíba: A Case Study of Family-Based Oligarchy
in Brazil. Princeton, NJ: Princeton University Press.
Lijphart, Arend. 1975. The Politics of Accommodation: Pluralism and Democracy in the Nether-
lands. 2nd ed. Berkeley: University of California Press.
Lindberg, Steffan I. 2003. ‘‘’It’s Our Time to Chop: Do Elections in Africa Feed Net-
Patrimonialism Rather than Counteract It?’’ Democratization 10, no. 2: 121–40.
Linz, Juan, J. 1990. ‘‘The Perils of Presidentialism.’’ Journal of Democracy 1, no. 1: 51–69.
references 325
Nelson, Joan M. 1989. Fragile Coalitions: The Politics of Economic Adjustment. U.S.–Third
World Policy Perspectives 12. New Brunswick, NJ: Transaction Books.
Nelson, Richard R., and Sidney G. Winter. 1982. An Evolutionary Theory of Economic Change.
Cambridge, MA: Harvard University Press.
Nicolau, Jairo M. 1997. ‘‘Notas sobre os Quatro Índices Mais Utilizados nos Estudos Elei-
torais.’’ In Olavo Brasil de Lima, Jr., ed., O Sistema Partidário Brasileiro. Rio de Janeiro,
Brazil: Fundação Getúlio Vargas.
Nieto, José Chanes. 1993. La designación del presidente de la República. Mexico City: Plaza y
Valdés Editors.
North, Douglass C. 1990. Institutions, Institutional Change, and Economic Performance. New
York: Cambridge University Press.
North, Douglass C., William Summerhill, and Barry R. Weingast. 2000. ‘‘Order, Disorder, and
Economic Change: Latin America versus North America.’’ In Bruce Bueno de Mesquita
and Hilton L. Root, eds., Governing for Prosperity. New Haven, CT: Yale University Press.
North, Douglass, and Barry Weingast. 1989. ‘‘Constitutions and Commitment: The Evolution
of Institutions Governing Public Choice in 17th-Century England.’’ Journal of Economic
History 49, no. 4: 803–32.
Ochoa García, Carlos. n.d. ‘‘Derechos Indígenas y pluralismo legal en América Latina.’’
Unpublished ms., Instituto Munik’, Guatemala City. Available at www.geocities.com/
alertanet.
O’Donnell, Guillermo. 1984. Y a Mí, Qué Me Importa? Notas Sobre Sociabilidad y Política en
Argentina y Brasil. Buenos Aires, Argentina: Cuadernos Center for the Study of State and
Society (English-language version in O’Donnell 1999a).
————. 1993. ‘‘On the State, Democratization, and Some Conceptual Problems: A Latin Ameri-
can View with Some Postcommunist Countries.’’ World Development 21, no. 8: 1355–69.
————. 1994. ‘‘Delegative Democracy.’’ Journal of Democracy 5, no. 1: 55–69.
————. 1996a. ‘‘Another Institutionalization: Latin America and Elsewhere.’’ Kellogg Institute
working paper 222. Notre Dame, IN: Kellogg Institute for International Studies. Mar.
————. 1996b. ‘‘Illusions about Consolidation.’’ Journal of Democracy 7, no. 2: 34–51.
————. 1999a. Counterpoints: Selected Essays on Authoritarianism and Democracy. Notre
Dame, IN: University of Notre Dame Press.
————. 1999b. ‘‘Horizontal Accountability in New Democracies.’’ In Andres Schedler, Larry
Diamond, and Marc C. Plattner, eds., The Self-Restraining State: Power and Accountability
in New Democracies. Boulder, CO: Lynne Reinner.
————. 1999c. ‘‘Polyarchies and the (Un)rule of Law in Latin America: A Partial Conclusion.’’
In Juan Méndez, Guillermo O’Donnell, and Paulo Sérgio Pinheiro, eds., The (Un)rule of
Law and the Underprivileged in Latin America. Notre Dame, IN: University of Notre Dame
Press.
O’Donnell, Guillermo, Jorge Vargas Cullel, and Osvaldo M. Iazzetta, eds. 2004. The Quality
of Democracy: Theory and Applications. Notre Dame, IN: University of Notre Dame Press.
Oiticica, Yulo. 2001. Dossiê Grupos de Extermínio. Salvador, Bahia, Brazil: Human Rights
Commission of the Legislative Assembly of Bahia.
Olukoshi, Adebeyo. 1998. ‘‘Introduction.’’ In Adebeyo Olukoshi, ed., The Politics of Opposition
in Contemporary Africa. Uppsala, Sweden: Afrika Institutet.
references 329
Olvera Aguirre, Carlos. 2001. ‘‘Fox Insta a los Yucatanos a Negociar en Tabasco?’’ Milenio
(Mexico City), Jan. 15.
Orellana Halkyer, René. 1998. ‘‘Escenarios locales de resolución de conflictos: Derechos
consuetudinario y recursos naturales en el Perú.’’ América Indígena 58, nos. 1–2:
223–35.
Otano, Rafael. 2001. ‘‘El nuevo partido transversal.’’ Chile Hoy, Jan. 6.
Ottaway, Marina. 2003. Democracy Challenged: The Rise of Semi-Authoritarianism. Wash-
ington, DC: Carnegie Endowment for International Peace.
Panebianco, Angelo. 1988. Political Parties: Organization and Power. Cambridge: Cambridge
University Press.
Paoli Bolio, Francisco José. 1985. ‘‘Legislación y proceso político, 1917-1982.’’ In Pablo Gon-
zález Casanova, ed., Las elecciones en México: Evolución y perspectivas. Mexico City: Siglo
Veintiuno.
Party of the Institutional Revolution. 1993. PRI: Documentos Básicos, 1993. Mexico City: Party
of the Institutional Revolution.
Pejovich, Svetozar. 1999. ‘‘The Effects of the Interaction of Formal and Informal Institutions
on Social Stability and Economic Development.’’ Journal of Markets and Morality 2, no. 2:
164–81.
Perelli, Carina. 1995. ‘‘La Personalización de la Política: Nuevos Caudillos, Outsiders, Política
Mediática y Política Informal.’’ In Carina Perelli, Sonia Picado, and Daniel Zovatto, eds.,
Partidos y Clase Política en América Latina en los 90. San José, Costa Rica: Instituto
Interamericano de Derechos Humanos.
Pérez Liñan, Aníbal. 2003. ‘‘Presidential Crises and Democratic Accountability in Latin Amer-
ica, 1990–1999.’’ In Susan Eva Eckstein and Timothy P. Wickham-Crowley, eds., What
Justice? Whose Justice? Fighting for Fairness in Latin America. Berkeley: University of
California Press.
Philip, George. 1992. The Presidency in Mexican Politics. New York: St. Martin’s Press.
Pierson, Paul. 2000. ‘‘Increasing Returns, Path Dependence, and the Study of Politics.’’ Ameri-
can Political Science Review 94, no. 2: 251–67.
————. 2004. Politics in Time: History, Institutions, and Social Analysis. Princeton, NJ: Prince-
ton University Press.
Pinheiro, Filho. 1998. Reforma Eleitoral: Voto Distrital Misto: A Soluçtao que o Brasil Quer
Cohnecer. Brasília, Brazil: Câmara dos Deputados.
Pitkin, Hanna F. 1967. The Concept of Representation. Berkeley: University of California Press.
Posas, Mario, and Rafael del Cid. 1983. La Construcción del Sector Público y del Estado
Nacional en Honduras, 1876–1979. 2nd ed. San José, Costa Rica: EDUCA.
Price, Robert M. 1975. Society and Bureaucracy in Contemporary Ghana. Berkeley: University
of California Press.
Putnam, Robert D. 1993. Making Democracy Work: Civic Traditions in Modern Italy. Prince-
ton, NJ: Princeton University Press.
Qué Pasa (Santiago). 2000. ‘‘Jaime Campos, fruto del cuoteo.’’ May 27.
————. 2002. ‘‘La tregua Lagos-Zaldívar.’’ Mar. 9.
Rabinowitz, George, and Stewart McDonald. 1989. ‘‘A Directional Theory of Issue Voting.’’
American Political Science Review 83: 93–121.
330 references
Siavelis, Peter. 1997a. ‘‘Continuity and Change in the Chilean Party System: On the Transfor-
mational Effects of Electoral Reform.’’ Comparative Political Studies 30, no. 6: 651–74.
————. 1997b. ‘‘Executive-Legislative Relations in Post-Pinochet Chile: A Preliminary Assess-
ment.’’ In Scott Mainwaring and Matthew Soberg Shugart, eds., Presidentialism and De-
mocracy in Latin America. New York: Cambridge University Press.
————. 2000. The President and Congress in Post-Authoritarian Chile: Institutional Constraints
to Democratic Consolidation. University Park: Pennsylvania State University Press.
————. 2002a. ‘‘Exaggerated Presidentialism and Moderate Presidents: Executive-Legislative
Relations in Chile.’’ In Scott Morgenstern and Benito Nacif, eds., Legislative Politics in
Latin America. Cambridge: Cambridge University Press.
————. 2002b. ‘‘The Hidden Logic of Candidate Selection for Chilean Parliamentary Elec-
tions.’’ Comparative Politics 34, no. 2: 419–38.
Sieder, Rachel. 1998. ‘‘Customary Law and Local Power in Guatemala.’’ In Rachel Sieder, ed.,
Guatemala after the Peace Accords. London: Institute of Latin American Studies.
————. 1999. ‘‘Rethinking Democratization and Citizenship: Legal Pluralism and Institutional
Reform in Guatemala.’’ Citizenship Studies 3, no. 1: 103–18.
————. 2003. ‘‘Renegotiating ‘Law and Order’: Judicial Reform and Citizen Responses in Post-
war Guatemala.’’ Democratization 10, no. 4: 137–60.
Sierra, María Teresa. 1998. ‘‘Autonomía y Pluralismo Jurídico: El Debate Mexicano.’’ América
Indígena 58, nos. 1–2: 21–39.
Sil, Rudra. 2001. ‘‘Privatization, Labor Politics, and the Firm in Post-Soviet Russia: Non-market
Norms, Market Institutions, and the Soviet Legacy.’’ In Christopher Candland and
Rudra Sil, eds., The Politics of Labor in a Global Age: Continuity and Change in Late-
Industrializing and Post-socialist Economies. New York: Oxford University Press.
————. 2002. Managing ‘‘Modernity’’: Work, Community, and Authority in Late-Industrializing
Japan and Russia. Ann Arbor: University of Michigan Press.
Silva, Eduardo. 1992. ‘‘Capitalist Regime Loyalties and Redemocratization in Chile.’’ Journal
of Interamerican Studies and World Affairs 34, no. 4: 77–117.
Skolnick, Jerome H., and James J. Fyfe. 1993. Above the Law: Police and the Excessive Use of
Force. New York: Free Press.
Smith, Peter. 1979. Labyrinths of Power: Political Recruitment in Twentieth Century Mexico
Princeton, NJ: Princeton University Press.
Snyder, Richard. 2001. ‘‘Scaling Down: The Subnational Comparative Method.’’ Studies in
Comparative International Development 36, no. 1: 93–110.
Sorauf, Frank. 1988. Money in American Elections. Glenview, IL: Scott, Foresman.
Souza, Celina. 1998. ‘‘Intermediação de Interesses Regionais no Brasil: O Impacto do Federa-
lismo e da Decentralização.’’ Dados 41, no. 3: 569–92.
Starn, Orin. 1992. ‘‘I Dreamed of Foxes and Hawks: Reflections on Peasant Protest, New Social
Movements, and the Rondas Campesinas of Northern Peru.’’ In Arturo Escobar and Sonia
E. Alvarez, eds., The Making of Social Movements in Latin America: Identity, Strategy and
Democracy. Boulder, CO: Westview Press.
————. 1999. Nightwatch: The Politics of Protest in the Andes. Durham, NC: Duke University
Press.
references 333
Stavenhagen, Rodolfo. 1988. Derecho Indígena y Derechos Humanos en América Latina. Mex-
ico City: Instituto Interamericano de Derechos Humanos, El Colegio de México.
Stokes, Susan C. 2001. Mandates and Democracies: Neoliberalism by Surprise in Latin Amer-
ica. New York: Cambridge University Press.
————. 2003. ‘‘Do Informal Rules Make Democracy Work? Accounting for Accountability in
Argentina.’’ Paper presented at the conference Informal Institutions in Latin America,
University of Notre Dame, IN, Apr. 23–24.
————. 2005. ‘‘Perverse Accountability: A Formal Model of Machine Politics with Evidence
from Argentina.’’ American Political Science Review 99 (Aug.): 315–25.
Sur, El (Concepción). 2003. ‘‘Cuanta cerda le queda a la Concertación.’’ July 20, p. 4.
Taylor, Michelle M. 1992. ‘‘Formal versus Informal Incentive Structures and Legislator Be-
havior: Evidence from Costa Rica.’’ Journal of Politics 54, no. 4: 1053–71.
————. 1996. ‘‘When Electoral and Party Institutions Interact to Produce Caudillo Politics: The
Case of Honduras.’’ Electoral Studies 15, no. 3: 327-37.
Taylor-Robinson, Michelle M., and Christopher Diaz. 1999. ‘‘Who Gets Legislation Passed in
a Marginal Legislature and Is the Label Marginal Legislature Still Appropriate? A Study of
the Honduran Congress.’’ Comparative Political Studies 32, no. 5: 590–626.
Teixeira, Tomaz, 1985. A Outra Face da Oliguarquia do Piauí (Depoimento). 2nd ed. Fortaleza,
Brazil: Stylus Comuncaçães Ltda.
Tercera, La (Santiago). 2000. ‘‘Presiones por los cupos de subsecretarios.’’ Feb. 1.
Thelen, Kathleen. 1999. ‘‘Historical Institutionalism in Comparative Politics.’’ Annual Review
of Political Science 2, no. 1: 369–404.
————. 2003. ‘‘How Institutions Evolve: Insights from Comparative Historical Analysis.’’ In
James Mahoney and Dietrich Rueschemeyer, eds., Comparative Historical Analysis in the
Social Sciences. New York: Cambridge University Press.
————. 2004. How Institutions Evolve: The Political Economy of Skills in Germany, Britain, the
United States, and Japan. New York: Cambridge University Press.
Tomz, Michael, Jason Wittenberg, and Gary King. 2003. Clarify: Software for Interpreting and
Presenting Statistical Results. Version 2.1. Stanford University, University of Wisconsin,
and Harvard University. Jan. 5. Available at http://gking.harvard.edu/clarify/docs/clarify
.html.
Torre, Juan Carlos. 1990. La vieja guardia sindical y Perón: sobre los orígenes del peronismo.
Buenos Aires, Argentina: Editorial Sudamericana.
Tsai, Kellee S. 2003. ‘‘Coping by Innovating: The Formal Origins and Consequences of
Informal Institutions in China.’’ Paper presented at the annual meeting of the American
Political Science Association, Philadelphia, Aug. 27–31.
Tsai, Lily Lee. 2001. ‘‘Substituting for the State? The Logic of Private Efficiency in Chinese
Local Governance.’’ Paper presented at the annual meeting of the American Political
Science Association, San Francisco, Aug. 30–Sept. 2.
————. 2002. ‘‘Cadres, Temple and Lineage Institutions, and Governance in Rural China.’’
China Journal 48 (July): 1–27.
————. 2004. ‘‘The Informal State: Governance and Development in Rural China.’’ Ph.D. diss.,
Department of Government, Harvard University, Cambridge, MA.
334 references
Ullman-Margalit, Edna. 1978. The Emergence of Norms. Oxford: Oxford University Press.
United Nations Development Program (UNDP). 2002. Hacia una integración cooperativa y
solidaria del territorio nacional. Buenos Aires, Argentina: United Nations Development
Program
Valencia Castrejón, Sergio. 1996. Poder regional y política nacional en México: El gobierno de
Maximino Avila Camacho en Puebla (1937–1941). Mexico City: Instituto Nacional de
Estudios Históricos de la Revolución Mexicana.
Valenzuela, Arturo 1994. ‘‘Party Politics and the Crisis of Presidentialism in Chile’’ In Juan
Linz and Arturo Valenzuela, eds., The Failure of Presidential Democracy, vol. 2. Baltimore:
Johns Hopkins University Press.
Valenzuela, J. Samuel, and Timothy Scully. 1997. ‘‘Electoral Choices and the Party System in
Chile: Continuities and Changes at the Recovery of Democracy.’’ Comparative Politics 29,
no. 4: 511–27.
Van Cott, Donna Lee. 2000a. The Friendly Liquidation of the Past: The Politics of Diversity in
Latin America. Pittsburgh: University of Pittsburgh Press.
————. 2000b. ‘‘A Political Analysis of Legal Pluralism in Bolivia and Colombia.’’ Journal of
Latin American Studies 32, no. 1: 207–34.
Vargas Manríquez, Fernando, and Guadalupe Moreno Corzo. 1994. ‘‘Legitimación Jurídica de
la Procuraduría Electoral del Pueblo Chiapaneco y el Valor Jurídico de sus Actuaciones y
Resoluciones.’’ Unpublished ms.
Von Benda-Beckmann, Franz. 1997. ‘‘Citizens, Strangers and Indigenous Peoples: Conceptual
Politics and Legal Pluralism.’’ Law and Anthropology 9: 1-42.
Walker, Ignacio. 2003. ‘‘Chile: Three Stories of Informal Institutions in a Limited Democ-
racy.’’ Paper presented at the conference Informal Institutions and Politics in Latin Amer-
ica, University of Notre Dame, South Bend, IN, Apr. 23–24.
Wang, Hongying. 2000. ‘‘Informal Institutions and Foreign Investment in China.’’ Pacific
Review 13, no. 4: 525–56.
Waterbury, John. 1973. ‘‘Endemic and Planned Corruption in a Monarchical Regime.’’ World
Politics 25, no. 4: 533–55.
Way, Lucan A. 2002. ‘‘The Dilemmas of Reform in Weak States: The Case of Post-Soviet Fiscal
Decentralization.’’ Politics and Society 30, no. 4: 579–98.
Wedeen, Lisa. 2003. ‘‘Seeing Like a Citizen, Acting Like a State: Exemplary Events in Unified
Yemen.’’ Comparative Studies in Society and History 45, no. 4: 680–713.
Weingast, Barry. 1979. ‘‘A Rational Choice Perspective on Congressional Norms.’’ American
Journal of Political Science 23, no. 2: 245–62.
Weingast, Barry R., and William J. Marshall. 1988. ‘‘The Industrial Organization of Congress;
or, Why Legislatures, Like Firms, Are Not Organized as Markets.’’ Journal of Political
Economy 96, no. 1: 132–63.
Weyland, Kurt. 1993. ‘‘The Rise and Fall of President Collor and Its Impact on Brazilian
Democracy.’’ Journal of Interamerican Studies and World Affairs 35, no. 1: 1–37.
————. 1996. ‘‘Neopopulism and Neoliberalism in Latin America.’’ Studies in Comparative
International Development 31, no. 1: 3–31.
————. 1997. ‘‘ ‘Growth with Equity’ in Chile’s New Democracy.’’ Latin American Research
Review 32, no. 1: 37–67.
references 335
————. 1999. ‘‘Neoliberal Populism in Latin America and Eastern Europe.’’ Comparative Poli-
tics 31, no. 4: 379–401.
————. 2002a. ‘‘Limitations of Rational-Choice Institutionalism for the Study of Latin Ameri-
can Politics.’’ Studies in Comparative International Development 37, no. 3: 57–85.
————. 2002b. The Politics of Market Reform in Fragile Democracies. Princeton, NJ: Princeton
University Press.
Whitehead, Laurence. 2002. Democratization: Theory and Experience. Oxford: Oxford Univer-
sity Press.
Wiarda, Howard, and Harvey Kline. 1996. Latin American Politics and Development Boulder,
CO: Westview Press.
Williamson, John. 1990. Latin American Adjustment: How Much Has Happened? Washington,
DC: Institute for International Economics.
Wright, Gerald C., Jr., and Michael B. Berkman. 1986. ‘‘Candidates and Policy in United
States Senate Elections.’’ American Political Science Review 80, no. 2: 567–88.
Yang, Mayfair Mei-hui. 1994. Gifts, Favors, and Banquets: The Art of Social Relationships in
China. Ithaca, NY: Cornell University Press.
Yashar, Deborah. 1999. ‘‘Democracy, Indigenous Movements, and the Postliberal Challenge
in Latin America.’’ World Politics 52, no. 1: 76–104.
Yrigoyen Fajardo, Raquel. 1999. Pautas de coordinación entre el derecho indígena y el derecho
estatal. Guatemala City: Fundación Myrna Mack.
————. 2000. ‘‘The Constitutional Recognition of Indigenous Law in Andean Countries.’’ In
Willem Assies, Gemma Van Der Haar, and Andre Hoekema, eds., The Challenge of
Diversity: Indigenous Peoples and Reform of the State in Latin America. Amsterdam: Thela
Thesis.
Yucatán State Legislature. 1993. Diario de Debates del Estado de Yucatán, 21 de diciembre,
1993. Mérida: Congresso Estatal de Yucatán, Mexico.
Zamarripa, Roberto. 1989. ‘‘Triunfo Perredista, según el Dictamen del Tribunal?’’ La Jornada
(Mexico City), Aug. 13.
Zorrilla, Ruben H. 1983. ‘‘Líder, Elite, y Masa en el Peronismo.’’ Todo Es Historia, Dec., pp.
28–37.
This page intentionally left blank
Contributors
editors
contributors
of Ethnic Politics (2005), The Friendly Liquidation of the Past: The Politics of
Diversity in Latin America (2000), and, as editor, Indigenous Peoples and
Democracy in Latin America (1994). Recent articles appeared in Comparative
Political Studies, Studies in Comparative International Development, and Latin
American Politics and Society.
Index
Italic page numbers indicate figures and tables. Argentina: accountability in, 128–39, 130, 132;
complementary institutions in, 14; economic
accommodating informal institutions, 15; cam- development in, 138; human rights laws in, 208;
paign finance contracts as, 89; in developing interpersonal trust in, 135–37, 136; party organi-
world, 278; and institutional stability, 17–18; zation in, 179, 182, 187–92, 195, 196; police kill-
judicial institutions as, 250, 311n2; relationship ings in, 201–2, 207–26; voters’ decision-rules in,
to formal institutions, 15, 17–18, 288–89 128–29, 131–34, 133, 134
accommodating informal institutions, Chilean, Assies, Willem, 273
33–55; cuoteo as, 36, 40–45; democracia de los authoritarian regimes: in developing world, 283,
acuerdos as, 36, 48–51; goals of, 39; and institu- 312n12; effects of informal institutions on, 11–12,
tional change, 18, 52; need for, 15, 35–36, 51–52; 283; executive turnover in, 143; in Mexico, 143,
origins of, 15, 51–54; partido transversal as, 36, 230–31
45–47 Autonomous Social Justice Commission
accountability, 125–39; in Argentina, 128–39, 130, (Ecuador), 253
132; definition of, 302n2; effects of informal Ávila Camacho, Manuel, 148, 150–51
institutions on, 9–11, 282; enforcement of, 128– Aylwin, Patricio, 45, 48–49, 175
29; expectations of, 128–39, 130, 132; horizontal Aymara communities, 261
vs. vertical, 9, 10; in institutional origins, 21; in
Mexico, 10, 229; party organization and, 196; Bachelet, Michelle, 175
regional variation in, 129–34, 130; responsive- Bartlett, Manuel, 158, 304n8
ness and, 302n2; roots of, 137–39; trust and, base units (UBs), 190, 192
135–37; in voters’ decision-rules, 126–27, 128 behavior patterns, vs. informal institutions, 6–7,
AD. See Democratic Action 204–5
Africa: informal institutions in, 276, 277, 281, 283; binomial system. See two-member district system
nondemocratic regimes in, 283, 312n12 blat, 278
Alemán, Miguel, 150–51, 303n3 Boeninger, Edgardo, 37, 45
Alianza por Chile, 33, 164, 174, 305n3 Bolivia: indigenous law in, 264, 265, 269; juntas
Allende, Salvador, 47, 48 vecinales in, 259–62
Alvarez, Luis H., 245 Borja, Rodrigo, 76
American Popular Revolutionary Alliance Borocz, Joszef, 276
(APRA) (Peru), 187, 258–59 bottom-up informal institutions, 18–19, 22
appointments, presidential: in Chile, 40–41, 166– Boylan, Delia, 48
74, 171, 172, 305n10; in Ecuador, 75, 76–77, 78, Brazil: campaign finance contracts in, 10, 95–104,
297n25; in Mexico, 150 96; clientelism in, 56–68; comparison of
APRA. See American Popular Revolutionary institutions within, 57–58; competing institu-
Alliance tions in, 16; federalism in, 56–57; formal elec-
342 index
Clouthier, Manuel, 233 solution of, possibility of, 55, 161–63, 177; doub-
coalitions: in Chile, 37–40, 53–54, 160–61; in ling by, 165–66, 305n5; electoral insurance in,
presidential vs. parliamentary systems, 160. See 161, 166–77; establishment of, 33, 160; partido
also ghost coalitions; specific coalitions transversal and, 45–47, 166–67; parties in, 160,
Cochabamba (Bolivia), 259–62 164; presidency lost by, possibility of, 174–76;
colaboracionismo, 295n9 presidential cooperation with, 37–40; social
collective-action problems: in Chile, 39, 47, 53– issues in, 161, 163; stability of, 33, 160; subpacts
54; in Mexico, 146–47 of, 42, 43, 54, 164; on two-member system,
Collier, Jane, 254 305n13
Collor de Mello, Fernando, 26, 102–3 Confederación de Producción y Comercio
Colombia, indigenous law in, 264, 265, 266, 269– (CPC) (Chile), 50
71 Congress (Brazil), campaign finance contracts in,
colonial law, 252, 277, 279 95–104
Colorados (Paraguay), 182 Congress (Chile): ballot for, 162, 163; cuoteo in,
Colosio, Luis Donaldo, 303n2 40–45; democracia de los acuerdos in, 48–51;
competing informal institutions, 15–16; in electoral insurance in, 167–68, 170–71, 173;
developing world, 276–77; institutional change majoritarianism in, 35, 42; majorities in, 37, 38,
and, 23; judicial institutions as, 250, 310n1, 161–63; partido transversal in, 46; presidential
311n2; party organizations as, 179; police killings cooperation with, 34, 37–40; weakness of, 33.
as, 16, 225–26; relationship to formal institu- See also two-member district system
tions, 15–16, 17, 288–89 Congress (Ecuador): agendas of, 76, 296n23; anti-
competition, electoral: and clientelism, 114, 115; gobiernista discourse in, 71–72; electoral system
and constituency services, 110; in dedazo, 24, for, 71, 76, 295n3, 296n22; impeachment by,
144, 147, 152–59; and demand for campaign 78–79, 79; party switching in, 77, 297n33,
funds, 92, 97–98; in two-member district sys- 298n44; presidential cooperation with, 69–70;
tem, 168, 177 presidents criticized by, 71–72, 295n7; pro-
complementary informal institutions, 13–14; in government coalitions in, 76, 296n24; public
developing world, 278–79, 280–81; electoral coalitions in, 70, 71–72; roll calls in, lack of, 70,
insurance as, 14, 176; ghost coalitions as, 74; 75. See also ghost coalitions
institutional change and, 17, 176; judicial Congress (Honduras): activities and duties in,
institutions as, 250, 311n2; relationship to formal 112–13; clientelism in, 106–24; locally targeted
institutions, 13–14, 288–89 bills in, 112–24, 116, 118–19; multiauthored bills
Concentration of Popular Forces (CFP) in, 300n10; parties dominating, 110–12, 300n12;
(Ecuador), 77, 196 reelection rates in, 114; smaller parties in, 111–
concertacesiones, in Mexico, 227–48, 236; 12, 300n7
accountability in, 10, 229; conditions necessary Congress (U.S.): campaign finance in, 94, 100;
for, 229, 239, 241, 247; definition of, 228; demise constituency service work in, 109; credit claim-
of, 228, 230, 238–45; electoral reform and, 229, ing in, 300n11; turnover in, 99
232, 234–40, 242; evolution of, 28, 228; implica- Conservative Party (PCE) (Ecuador), 81
tions for democratization, 12, 245–48; institu- consociationalism, 15
tional change and, 23, 24; need for, 227–28; constituency services: electoral systems and, 109;
origins of, 20, 227–28, 230–34; presidential elec- in Honduras, 112–24
tions and, 227, 229, 232, 233–34, 236, 241; as sub- constitution(s): of Argentina, 208; of Brazil, 208;
stitutive institution, 16, 228, 235, 245–48, 292n26 of Chile, 15, 18, 52; of Colombia, 266; indige-
Concertación coalition (Chile), 33; accommodat- nous law in, 264, 265, 267–68, 269; of Mexico,
ing institutions built by, 15, 36; complementary 143, 144, 145, 148–49, 159, 303n3; of Peru, 259;
institutions and, 14; cuoteo and, 40–45; democ- presidential power in, 1, 2; of U.S., 14
racia de los acuerdos and, 48–51, 294n15; dis- Constitutional Court (Colombia), 269–71
344 index
Córdoba (Argentina): accountability in, 130, 130– Diaz Ordaz, Gustavo, 155
34; interpersonal trust in, 135–37, 136; police district magnitude (DM), 108, 112, 114, 115–17
killings in, 207–26, 211, 212 doubling, 164, 164–66
corporal punishment, 256–57, 268, 311n4 due process, 268, 270
corporate contributions, 100–101, 102–3 Durán Ballén, Sixto, 70, 79–83
Correa, Enrique, 45, 47
corruption: in Brazil, 102–3; clientelism associ- economic crises: in Ecuador, 71, 295n2; in
ated with, 296n13; as competing institution, 15– Mexico, 152–53, 154, 155
16; in developing world, 276; in Ecuador, 72, economic development: and accountability, 138;
80, 82–83; graft, 7; as informal behavior vs. in Brazil, 58, 61, 101; in Honduras, 111
institution, 7; institutional change and, 23 economic reform: in Ecuador, 69–73, 80–83; in
Costa Rica, 106–7, 311n8 Mexico, 234–35
Council against Police and Institutional Ecuador: formal electoral system of, 71, 76, 295n3,
Repression, 210 296n22; formal institutions of, 70–73; ghost
CPC. See Confederación de Producción y coalitions in, 69–84; identification of institu-
Comercio tions in, 26–27; indigenous law in, 253, 264,
credit claiming, legislative, 112, 300n11 265, 269; institutional change in, 23; modern-
culture, vs. informal institutions, 5, 7–8, 207 ization reforms in, 70, 73, 80-83; party organiza-
cuoteo, in Chile, 36, 40–45, 294n3 tion in, 179, 187, 192–95, 196
customary law, 251, 264, 267 Ecuadorian Roldosista Party (PRE), 77, 80, 179,
187, 192–95
Dahik, Alberto, 27, 80, 81, 82–83 education: and accountability, 138; in Brazil, 61,
dedazo, in Mexico, 143–59; accountability in, 10; 62; and clientelism, 61, 114–15; in Honduras,
collapse of, 24, 144–45, 152–58; comparison of 114–15, 117–21
cases of, 27–28; definition of, 1, 143; electoral effectiveness, of formal institutions: vs. efficiency,
competition and, 24, 144, 147, 152–59; enforce- 292n16; and institutional change, 23–24; in
ment of, 144, 145; formal institutions and, 151– typology, 13–16, 88
52, 154–59, 303n7; governance in, 11–12; institu- efficiency: accountability and, 130–32; vs. effec-
tionalization of, 144, 145, 149–52; nomination tiveness, 292n16
process in, 145, 146, 156–58, 304n11; origins of, El Alto (Bolivia), 259–62
20, 21, 144, 147–49, 293n34; public mobilization Electoral College (Mexico), 234, 235–38
in, 146, 150; regime ruptures challenging, 149– electoral courts, in Mexico, 235–37, 238, 241–45
51, 152–53; selection process in, 143–44, 145–46, electoral insurance, in Chile, 161, 166–77;
154–56 appointments in, 166–74, 171, 172; as comple-
Degregori, Carlos Iván, 259 mentary institution, 14, 176; cuoteo and, 44; and
Delfim Neto, Antônio, 101 governance, 11; institutional change and, 23, 24,
democracia de los acuerdos, in Chile, 36, 48–51 176–77; origins of, 20; threats to future of, 174–
Democratic Action (AD) (Venezuela), 181, 185, 76
186 electoral markets: clientelistic, 56–68; as informal
Democratic Current (CD) (Mexico), 153 institutions, 59–61; programmatic, 58, 60–68,
Democratic Left (ID) (Ecuador), 71–72, 296n16 65. See also clientelism
Democratic Party (PD) (Ecuador), 74 electoral reform, in Mexico, 229, 232, 234–40, 242
Democratic Unification (PUD) (Honduras), 112 electoral systems, formal: accountability in, 10, 14;
demonstrations, and concertacesiones, 231–33, of Brazil, 56–57, 92, 98; of Chile, 35, 42, 57; of
238–39 Ecuador, 71, 76, 295n3, 296n22; literature on,
D’hondt method, 164 107–10; of Mexico, 151–52, 154–59, 227–28, 237,
Dia, Mamadou, 279, 281 303n7; personal vs. partisan voting in, 108, 112,
Diaz, Porfirio, 148 121–22. See also specific types
index 345
Electoral Tribunal of the Judicial Power of the Ferreira dos Santos, Manuel, 220–21, 223
Federation (TEPJF) (Mexico), 238, 240, 243 finance, in party organizations, 186, 191–92, 195.
elites: accommodating institutions built by, 35– See also campaign finance
36, 51–54, 55; in cuoteo, 40, 41–42, 44, 45; in Fiscal Responsibility, Stabilization and Trans-
dedazo, 143–44, 146–49; in democracia de los parency Law (Ecuador, 2002), 73
acuerdos, 49–50; in institutional origins, 18, 22; foot-binding, 25
motivations of, 51–52; in partido transversal, 36, formal institutions: change in, 17–18, 23–24; con-
166 vergence in outcomes of, 13–16, 89; definition
Ellickson, Robert C., 5, 204, 205 of, 5; effectiveness of, 13–16, 23–24; origins of,
enforcement: of accommodating institutions, 53– 19; shifts between informal and, 286–87; in
54; of accountability, 128–29; of campaign typology, 13–19
finance contracts, 20, 87, 88, 92–95, 101–4; of Fox, Vicente, 227, 241, 242, 243, 245, 310n22
clientelism, 59, 127–28; of complementary fraud, electoral, in Mexico, 232, 234, 242
institutions, 14; of cuoteo, 41, 43–44; of dedazo, Frei, Eduardo, 41, 175
144, 145; of democracia de los acuerdos, 50; of Fujimori, Alberto, 182
ghost coalitions, 26–27, 70, 74, 78–80; in iden- functionality: of democracy, 11–12, 129–39, 196,
tification of institutions, 26–27, 203–8, 206; of 275; of informal institutions, 16–17
indigenous law, 268, 270–71; internal vs. exter- fundraising, 186
nal, 5–6, 291n7; of juntas vecinales, 260–61; of
partido transversal, 47; of police killings, 215– Galvan, Dennis, 280, 281
21, 224; of programmatic electoral markets, 60; García, Alan, 259
of rondas campesinas, 256–57, 311n4; of voters’ García, Roberto, 187, 189
decision-rules, 127–28 General Assembly of Neighbors, 260
equality: and accountability, 138; in rule of law, General Index of Structural Reform, 73, 295n11
201 gentleman’s agreements. See concertacesiones
Esparza, Gary, 74, 296n16, 296n17 ghost coalitions, in Ecuador, 69–84; clientelism
ethnography, 25, 26, 203 in, 73–74; as complementary institution, 74;
executive branch: abuses by, 6; checks on, in rule definition of, 28, 73; under Durán Ballén, 79–
of law, 201; Chilean accommodating institu- 83; in economic reform, 69–70, 80–83;
tions in, 40–41, 46, 48–50; in clientelistic vs. enforcement of, 26–27, 70, 74, 78–80; formal
programmatic markets, 60–61; effects of infor- context for, 70–73; formation of, 73–80; identi-
mal institutions on, 11. See also presidents fication of, 26–27; and modernization reforms,
expectations, shared: of accountability, 128–39; in 70, 80–83; patronage in, 75–78; political
identification of institutions, 26, 204; vs. infor- reform and, 84, 298nn48&49; proto-parties in,
mal rules, 288; and institutional change, 24–25; 77–78; revelation of, threat of, 80, 82, 83; trade-
normative, 204; vs. shared values, 7–8, 204; of offs of, 75–76, 84; vote-buying in, 11, 70, 74, 75–
voters’ decision-rules, 131–34, 133, 134 78; written agreements on, 79–80, 81–82,
297n39
Farias, Paulo César, 102–3 gift-giving norms, 278
favelas, 63 gobernación, secretary of (Interior Secretariat)
Fearon, James, 134 (Mexico), 146, 151, 235, 237, 244
Febres Cordero, Léon, 79, 81, 82, 297n38 governors, Brazilian: in clientelism, 59, 63; and
Federal Electoral Commission (Mexico), 237 party cohesion, 64, 65–67; in programmatic
Federal Electoral Institute (IFE) (Mexico), 237, markets, 61
240 governors, Ecuadorian, 76, 78
Federal Electoral Tribunal (Mexico), 237 governors, Mexican: in concertacesiones, 231, 233,
Federation of Juntas Vecinales, 260 236, 238–39, 241, 242; in dedazo, 155, 156–57,
Fernández de Cevallos, Diego, 235 304n8; in municipal councils, 231, 309n4
346 index
GR. See Roldosista Group Innovation and National Unity Party (PINU)
grupos de extermínio, 307n6 (Honduras), 111–12, 113
Grzymala-Busse, Anna, 279, 311n5 institutions: definition of, 5; formal vs. informal,
Guanajuato (Mexico), 245–46 5–6; implicit vs. explicit understanding of, 125–
guanxi, 278 26. See also specific types
Guaraní people, 255 Interior Secretariat. See gobernación, secretary of
Guatemala: indigenous law in, 264, 265, 271, 272; International Labor Organization (ILO) Conven-
institutional change in, 24 tion 169, 264–66
Italy, 15–16, 23, 137
hacienda justice, 256, 257
Hamilton-Hart, Natasha, 278–79 Jalisco (Mexico), 241–42
Henriqúez, Miguel, 161 Jambaló (Colombia), 270–71
Honduras: clientelism in, 106–24; local legisla- judicial institutions, formal: police killings toler-
tion, 112-24, 116, 118-19 ated by, 205, 208–26; weakness of, 249. See also
Huejotzingo (Mexico), 239–40 specific types
Huerta, Adolfo De la, 147 judicial institutions, informal, 249–73; com-
human rights laws, 20, 208 parison of, 262–63, 263; formalization of, 266–
Hungary, 279 67; formal recognition of, 250, 264–72; indige-
Hurtado, Osvaldo, 74, 76, 77–78 nous law as, 251–55; juntas vecinales as, 259–62;
rondas campesinas as, 255–59; state relation-
ID. See Democratic Left ships to, 250, 264–72; in typology, 250, 310n1,
IFE. See Federal Electoral Institute 311n2
ILO. See International Labor Organization juntas vecinales, in Bolivia, 259–63, 263
impeachment: campaign finance and, 102–3; jurisprudence, and indigenous law, 269–71
ghost coalitions and, 78–79, 79, 82 Jusicialista (Peronist) Party (PJ) (Argentina), 179,
Independent Democratic Union (UDI) (Chile), 182, 187–92, 195, 196
173–74 Justice and Peace Commission of the Archdiocese
indigenous law, 251–55; colonial law and, 252, 277, of Salvador, 209
279; vs. culture, 7; definition of, 251; in develop- justiceiros, 214
ing world, 277, 281; evolution of, 251, 252–53; justices of the peace, 271–72
formalization of, 266–67; formal recognition
of, 264–72, 265; as formal vs. informal, 266–67, Kirchner, Néstor, 187–89
289; origins of, 7, 252; vs. other judicial systems, Knight, Alan, 147
262–63, 263; state relationship to, 253–55, 264– Knight, Jack, 5, 22, 24
72 Krauss, Enrique, 45
informal institutions: ambiguity of concept, 4; Kymlicka, Will, 269, 271
change in, 22–25, 275; common knowledge of,
286; comparisons of, 27–28, 57–58; conver- Labastida, Francisco, 158
gence in outcomes of, 13–16, 89; definitions of, labor unions, 191
1, 4–8, 87, 202–8, 286; vs. expectations, 288; Lagos, Ricardo, 174–75
identification of, 25–28, 203–8, 206; influence Lambayeque (Peru), 257
of, 1–3, 8–13; operationalization of, 202–8; La Paz (Bolivia), 259–62
origins of, 18–21; vs. other informal phe- Lauth, Hans-Joachim, 4, 22
nomena, 6–8; shifts between formal and, 286– Lavín, Joaquín, 174–75
87; significance of, 8–13; typology of, 13–19, 14, legal pluralism, 250, 264, 270, 273, 310n1
88–89, 250, 288. See also specific types legislatures: credit claiming in, 112, 300n11; repre-
informal organizations, 7 sentation as function of, 106–7. See also Con-
informal rules, definition of, 1, 202–8, 288 gress; state legislatures
index 347
Liberal Front Party (PFL) (Brazil), 65–66 M=2 district system. See two-member (M=2)
Liberal Party (PLH) (Honduras), 110–11, 113, 121 district system
Libro de Actas, 260, 261, 262 multiculturalism, 264, 267, 271, 310n1
Linz, Juan, 280 municipal councils, in Mexico, 231–32, 309n4
‘‘little bird’’ torture, 256, 311n4 Muñoz Ledo, Porfirio, 303n6
López Obrador, Andrés Manuel, 238
Los Angeles (California), 209 Nader, Laura, 273
Lozano, Rodríguez, 235 National Action Party (PAN) (Mexico): in concer-
Lucero, Wilfrido, 71–72 tacesiones, 10, 227–48, 309n2; on electoral
Lula (Luiz Inácio da Silva), 102 reform, 242, 310n19; electoral success of, 230,
232; in 1976 presidential election, 229, 232; orga-
machine politics, 109–10, 182, 185, 188, 195 nization of, 181, 185, 186
Mackie, Gerry, 25 National Assembly, of PRI, 155–56, 157
Madrazo, Roberto, 158, 239, 240, 242, 304n8 National Concentration Legislative Bloc, 79–80,
Madrid, Miguel de la, 152–53, 303n6 81–82
Mahuad, Jamil, 71–72, 74–75, 295n7 National Executive Committee of PRI (CEN)
Mainwaring, Scott, 296n13 (Mexico), 155
majoritarianism, in Chile, 35, 42 National Integration Movement (MIN), 75, 78
majority parties: in Chile, 37, 38, 161–63; in National Party (PNH) (Honduras), 110–11, 113, 121
Ecuador, 76, 296n24; in Honduras, 110–12, National Political Council (Mexico), 157, 158
300n12 Nebot, Jaime, 81, 83
March, James, 52 neoliberalism, 234–35
marginal list position, 114, 117 neopatrimonialism. See patrimonialism
Mayan communities, 272 Nicaragua, indigenous law in, 264, 265
Mayhew, David, 300n11 Noboa, Gustavo, 75, 77, 78, 297n31
mayors, in concertacesiones, 231, 232, 235–37, 236 norms, informal: in developing world, 278;
Méndez, Juan E., 202 enforcement of, 5, 291n7; of restraint, 282–83;
Menem, Carlos, 17, 187, 191 societal, 5; use of term, 204, 291n7
Mershon, Carol A., 52, 297n26 North, Douglass, 1, 7, 95
Mexico: authoritarian regime in, 143, 230–31; con-
certacesiones in, 227–48; dedazo in, 143–59; Obregón, Álvaro, 147, 303n3
economic crises in, 152–53, 154, 155; indigenous O’Donnell, Guillermo, 1, 2, 6, 9, 12, 30, 201, 202,
law in, 264, 265; institutional change in, 23–24; 249, 272
origins of institutions in, 20, 21; substitutive Olsen, Johan, 52
institutions in, 16 open-list proportional representation, 57, 64, 92,
Mexico City, 209 98, 108, 109
Michoacán (Mexico), 233 opposition parties: cohesion in, 64–67, 65; in con-
MIN. See National Integration Movement certacesiones, 229, 231–33; in dedazo, 153; in
Ministry of the General Presidency (SEGPRES) democracia de los acuerdos, 48–50; local legisla-
(Chile), 45–46 tion by, 113, 120, 300n12. See also ghost coali-
Misiones (Argentina), 130, 130–34, 135–37, 136 tions; specific parties
Molinar, Juan, 151 organization rules, 5
Monetary Stability and Economic Recovery Law
(Ecuador, 2000), 73, 295n10 PACs. See political action committees
Montero, Jorge, 80, 298n40 Páez Cabildo Association of the North, 270–71
Montevideo (Uruguay), 207–8 PAN. See National Action Party
Montonero guerrillas, 190 Panama, indigenous law in, 264, 265
Moore, Sally Falk, 273 Panebianco, Angelo, 180
348 index
paraconstitutional rules, 3, 280 PINU. See Innovation and National Unity Party
Paraguay, indigenous law in, 264, 265 Piura (Peru), 257
‘‘parchment’’ institutions, 1, 274 PJ. See Jusicialista (Peronist) Party
Paredes, Orlando, 235, 309n11 PLH. See Liberal Party
particularism: and accountability, 10; as compet- pluralism, legal, 250, 264, 270, 273, 310n1
ing institution, 16; definition of, 276; in PMDB. See Brazilian Democratic Movement
developing world, 276; and representation, 9 Party
partido transversal, in Chile, 45–47; definition of, PNH. See National Party
36, 45, 166; in democracia de los acuerdos, 48; in police killings, 201–26; case studies of, 27, 216–22;
electoral insurance, 166–67 conviction rates for, 202, 210–13, 211, 212; data
Party of the Democratic Revolution (PRD) on, 307n8; enforcement of rule of, 215–21, 224;
(Mexico), 229, 233, 234, 238–39, 309n2 evidence in, 218-19; as informal institution, 16,
Party of the Institutional Revolution (PRI) (Mex- 225–26; laws against, 20, 208; media coverage
ico): authoritarian regime under, 143, 230–31; of, 214, 217; origins of, 21, 202; permissive rules
in concertacesiones, 30, 227–48; in dedazo, 143– in, 205, 210; public opinion of, 213–14
59; domination of, 232, 303n1; enforcement by, political action committees (PACs), 100
27; establishment of, 303n1; informal institu- political parties: accountability in, 9–10; brand
tions of, 143; National Assembly of, 155–56; names for, 108; in clientelism, 59, 61; definition
organization of, 182, 185, 187; platform of, 230– of, 180; in developing world, 277; governance
31; sectors of, 145, 146 in, 11; institutionalization of, 180; in institu-
party system: of Brazil, 57, 97; of Chile, 11, 35, 37, tional origins, 21; partisan voting for, 108; repre-
47, 55, 160, 164, 177; of Ecuador, 11, 69–70, 83 sentation by, 8–9, 106–7; switching between,
patrimonialism and neopatrimonialism, 2, 8, 16, 77, 297n33, 298n44. See also coalitions; opposi-
34, 275, 276 tion parties; specific parties
patronage, in ghost coalitions, 73–74, 75–78. See political parties, Brazilian: and campaign finance,
also clientelism 92, 97–98; in clientelistic vs. programmatic
PCCh. See Chilean Communist Party markets, 59, 61, 64–67, 294n2; cohesion of, 64–
PCE. See Conservative Party 67, 65; Collor’s attitude toward, 103; electoral
PD. See Democratic Party system’s influence on, 57, 92, 98
PDCH. See Christian Democratic Party political parties, Chilean: in cuoteo, 40–45; in
Peace and Justice Service, 210 democracia de los acuerdos, 48–51; electoral sys-
Pejovich, Svetozar, 5 tem’s influence on, 35; in partido transversal,
people’s courts, in Mexico, 232–33, 243–44 45–47; small, 40, 43, 54; subpacts of, 42, 43–44,
permissive rules, 27, 205, 210 54
Perón, Isabel, 191 political parties, Costa Rican, 106–7
Perón, Juan, 187, 190 political parties, Ecuadorian, 71, 77, 297n33,
Peronist Party. See Jusicialista (Peronist) Party 298n44
Peru: indigenous law in, 264, 265, 269, 271–72; political parties, Honduran: in clientelism, 110–11,
institutional change in, 24; origins of institu- 123–24; local legislation by, 112–24; small, 111–
tions in, 20; rondas campesinas in, 255–59, 262– 12, 300n7
63, 271–72; substitutive institutions in, 16 political parties, Mexican: in concertacesiones,
PFL. See Liberal Front Party 227–48; in dedazo, 143–59
Piauí (Brazil): electoral markets in, 58, 61–68; political parties, Peruvian, 258–59
party cohesion in, 64–67, 65; poverty in, 58, political party organizations, 178–97; ancillary
61–62, 62 organizations of, 185–86, 191, 194; in Argentina,
Piñera, Sebastián, 175 179, 182, 187–92, 195, 196; boundaries of, 184,
Pinheiro, Paulo Sérgio, 202 190, 194; bureaucracy of, 181, 182–83, 189–90,
Pinochet, Augusto, 160, 161 193; career paths in, 184–85, 190–91, 193–94;
index 349
United States: campaign finance in, 94, 95, 100; tations of, 131–34, 133, 134; grammatical, 3, 125–
complementary institutions in, 14; institutional 28, 302n1; retrospective egocentric, 126–27;
change in, 23; institutional origins in, 20; police retrospective sociotropic, 126–28; socialization,
killings in, 209; presidentialism in, 280–81; pri- 126; spatial, 126–27; strategic, 126–27
maries in, 195–96 voting roll call: in Brazil, 60, 63, 64–67, 65; in
U.S. House of Representatives, 94, 99, 293n41 Ecuador, 70, 75
Uruguay: human rights laws in, 208; police kill-
ings in, 207–26, 211, 212 Walker, Ignacio, 45, 46
Wang, Hongying, 277–78
Vara Criminal Especializada da Infância e da weak institutions: in developing world, 279–80;
Juventude (Brazil), 222–23 vs. informal institutions, 6; in origins of institu-
Velasquez, Fidel, 150 tions, 20; party organizations as, 178, 180–81;
Venezuela, indigenous law in, 264, 265, 269 substitutive institutions and, 16
veto power, 49–50, 71 Wedeen, Lisa, 277
‘‘Villa Ramallo’’ case, 217 Weingast, Barry, 95
violence: in rondas campesinas, 256–57, 311n5; by Weyland, Kurt, 39, 53
victims of police killings, 211–21, 212 Worker’s Party (PT) (Brazil), 66, 67, 181, 186
vote-buying, legislative, in Ecuador, 11, 70, 74,
75–78 Yrigoyen Fajardo, Raquel, 273
voters: clientelism and, 112; in concertacesiones,
228; personal vs. partisan, 108, 112, 121–22 Zapatista rebels, 154
voting decision-rules, 126; clientelistic, 126–28, Zedillo, Ernesto, 154–58, 234, 238–40, 304n8
133, 134; directional, 126–27; ethnic, 126; expec-